一种改进的变步长自适应GSO算法

基本萤火虫群优化(GSO)算法在求解全局优化问题时，存在收敛速度慢、求解精度不高等问题。为此，提出一种变步长自适应GSO算法。该算法在一定程度上可以避免GSO算法过早陷入局部最优，且步长随迭代次数的增加而自适应地调整，从而使算法在后期获得精度更高的解。运用6个标准测试函数进行实验，结果表明，与GSO算法相比，该算法的收敛速度及精度均有明显提高。

Improved Variation Step Adaptive GSO Algorithm

An improved Variation Step Adaptive Glowworm Swarm Optimization（VSAGSO） algorithm is proposed to solve the problem of slow convergence and low precision and easy to fall into local optimization of the Glowworm Swarm Optimization（GSO） algorithm. It endows a big initial step to each glowworm. The step is decreased dynamically along with the increase of iteration so that the algorithm can get more precise solution in the end of the algorithm. Experimental results with six test function show convergence speed and precision is dramatically improved, which testifies that VSAGSO is a valid method to solve the global optimization problem.