The Research Of Explosion Search Algorithm

Now more and more practical engineering problems'characteristics present complex, multipolar, non-liner, strong restriction and so on, so that they can't be solved by the traditional algorithm, which drives people to seek more efficient optimization algorithm. Getting enlightenment from characteristics of life, many intelligent optimization algorithm are given to solve these complex problems. However, each new algorithm has some shortages in it need improving. new algorithm is created in this paper, based on the idea of bombing, whose name is Explosion Search Algorithm.In this paper, several existing intelligent optimization algorithms are analysed firstly with their advantages and shortcomings. After that, one new algorithm, Explosion Search Algorithm(ESA), is given, which is composed with three core arithmetic operators: Explosion Search Operator, Moving Operator, Mutation Operator. One theory defined as Neighborhood Search is proposed in ESA. This new algorithm has stronger ability of global search as well as local search. Both parallel global search and parallel local search are used in ESA, which can improve the search capability of it. An implementation method to realize the algorithm is given after it in the paper. The simulation using 20 standard benchmark functions and comparison with other algorithms proved the efficiency of the new algorithm.There still are some shortages in ESA needing improving. Further improvements to the shortages shown from the simulation experiment are implemented. Firstly we improve the Explosion Search Operator to Reduce the time complexity. Secondly, the idea of steepest descent is Introduced into the ESA by adding new operator, so that the accuracy of the algorithm is Improved. The improved ESA reduces the time complexity, increase the search accuracy Compared with the ESA. The simulation of Benchmark functions can Verify it.There are some advantages, such as good searching ability, fast convergence, high precision and High stability, in the improved ESA. So this algorithm Absolutely is an efficient algorithm. Future, the ESA can be combined with the practical problems to make further optimized.