Community structure is one of the most important attributes of complex networks, which reveals the hidden rules and behavior characteristics of complex networks. Existing works need to pre-set weight parameters to control the different emphasis on the objective function, and cannot automatically identify the number of communities. In the process of optimization, there will be some challenges, such as premature and inefficiency. This paper presents a multi-objective adaptive fast evolutionary algorithm (F-SGCD) for community detection in complex networks. Firstly, it transforms the problem of community detection into a multi-objective optimization problem and constructs two objective functions of community score and community fitness. Secondly, an external elite gene pool is introduced to store non-inferior solutions with high fitness. At the same time, an adaptive genetic operator is executed to return a set of non-dominant solutions compromised between the two objective functions. Finally, a Pareto optimal solution with the highest modularity is selected and decoded to generate a set of independent subnetworks. Experiments show that the multi-objective adaptive fast evolutionary algorithm greatly improves the accuracy of community detection in complex networks, and can discover the hierarchical structure of complex networks better.

社区结构是复杂网络最重要的属性之一，它揭示了复杂网络的隐藏规则和行为特征。现有作品需要预先设置权重参数，以控制对目标函数的不同重视，并且无法自动识别社区数量。在优化过程中，会遇到一些挑战，例如过早和效率低下。本文提出了一种用于复杂网络社区检测的多目标自适应快速进化算法（F-SGCD）。首先，将社区检测问题转化为多目标优化问题，并构建了社区评分和社区适应度两个目标函数。其次，引入了外部精英基因库来存储具有高度适应性的非劣等解决方案。同时，执行自适应遗传算子以返回在两个目标函数之间折衷的一组非主要解。最后，选择并解码具有最高模块化的Pareto最优解决方案，以生成一组独立的子网。实验表明，多目标自适应快速进化算法大大提高了复杂网络中社区检测的准确性，可以更好地发现复杂网络的层次结构。