Orthogonal Chemical Reaction Optimization Algorithm For Community Detection In Complex Networks

With the development of complex network research,people gradually began to pay attention to the complexity of network structure and its relationship with network behavior.In order to better understand the relationship between network structure and network behavior,it is necessary to understand in detail the characteristics of the network.Complex network has many characteristics.Community structure is one of the important properties.Its in-depth study plays an important role in the research of complex networks.The community structure is usually expressed as a set of nodes in the network.The entire network is composed of multiple clusters.The connections between the internal nodes of each community are relatively close,and the connections between various communities are sparse.Some features of the network are closely related to the community structure.Analyzing the community structure of the network can better understand the relationship between network structure and network behavior.This thesis focuses on the problem of community structure detection in complex networks and designs algorithms that are suitable for discovering hidden community structures in complex networks.The main tasks and innovations are as follows:1.A community detection method based on orthogonal chemical reaction optimization algorithm was proposed.The proposed algorithm converts the detection problem of the community structure into an optimization problem for solving the lowest potential energy.By selecting the optimization function,the nature and characteristics of the community structure are analyzed,and four basic chemical reaction operators suitable for the community detection problem are designed.In order to avoid the premature convergence of the algorithm and promote the uniformity of individual's distribution,this thesis improves the synthesis reaction operator based on the orthogonal experiment design.The orthogonal cross operator is used as the operator of the synthesis reaction.Its characteristic is the use of orthogonal tables.The creation of a new molecular entity after the synthesis reaction increases the quality of the individual solution of the synthetic molecule.2.Combine the chemical reaction optimization algorithm and the simulated annealing algorithm,and propose a local search strategy.By using the simulated annealing algorithm as a local search operator to improve the local search ability of the algorithm,it can avoid the algorithm falling into the local Local optimal solution.Experiments and analysis show that the proposed algorithm has strong feasibility and high efficiency in artificially synthesized network and real network data.Compared with traditional detection methods,it has significant advantages in overcoming the limitation of resolution.3.In order to make the performance of the experimental results more intuitive,this thesis designs a network visualization interface based on the Matlab GUI development platform.In the design process,this thesis uses the Archimedes spiral coordinate equation to lay out the nodes in the network,and distinguishes the communities to which each node belongs by coloring these nodes.Finally,In order to more easily observe the division results of the network community structure,we draw a topology diagram of the entire network.