1: \begin{abstract}

2: This paper presents a cooperative framework for fireworks algorithm (CoFFWA).

3: %Based on the analyses, the following observations are made:

4: A detailed analysis of existing fireworks algorithm (FWA) and its recently developed variants has revealed that ($i$) the selection strategy lead to

5: the contribution of the \textit{firework with the best fitness} (core firework) for the optimization overwhelms the contributions of \textit{the rest of fireworks} (non-core fireworks) in the explosion operator, ($ii$) the Gaussian mutation operator is not as effective as it is designed to be.

6: %

7: To overcome these limitations,

8: the CoFFWA is proposed, which can greatly enhance the exploitation ability of non-core fireworks by using independent selection operator and increase the exploration capacity by crowdness-avoiding cooperative strategy among the fireworks.

9: %

10: Experimental results on the CEC2013 benchmark functions suggest that CoFFWA outperforms the state-of-the-art FWA variants, artificial

11: bee colony, differential evolution,  the standard particle swarm optimization (SPSO) in 2007 and the most recent SPSO in 2011 in term of convergence performance.

12: \end{abstract}

13: