An improved Gravitational Search Algorithm based on single dimension swimming

Gravitational Search Algorithm (GSA) is a novel intelligent optimization algorithm that has high search ability and fast convergence. However, the standard GSA is easy to fall into local optimum and has low solution precision for complex optimization problems. To overcome these drawbacks, an improved Gravitational Search Algorithm based on Single Dimension Swimming (SDSGSA) is proposed. First, chaos sequence is adopted to initialize the population. Then, a new motion mode of single dimension swimming is proposed. Finally, mutation based on t-distribution is applied to the first k agents with the best fitness in each iteration. Simulation experiments on ten standard benchmark functions are carried out, and the results show that the proposed algorithm has high solution precision and fast convergence without premature convergence.