耦合体系拓扑空间的局部损坏与子空间的重构

摘要

In this paper,to investigate the effect of the local change or defect in the topological structure on finding the optimal path,it is assumed that there are a number of possible paths to form a network topology between nest and food source.To calculate the corresponding connection channel,the ant colony optimization algorithm is used,and the corresponding topological structure is constructed w hen the transfer coupling system reaches the optimal state.It is found that a new path channel could be searched out,on the one hand,the ant colony can find out the alternative path,so the function of the system could be restored and the effects of local defects could be compensated by constructing the subspace.On the other hand,the best path of ant searching is related to the length of time of local damage,which corresponds to the local optimal solution and the combinatorial optimal solution respectively.Our research indicates that by using ant colony optimization algorithm,the information channel could be repaired by constructing a new subspace.Furthermore,by analysing the phenomenon of information overload,we can forecast and check the local faults in the neural network or information network.%本文设想由蚁穴和食物源之间多条可能路径形成一个网络结构空间,考察该空间的拓扑结构发生局部改变或出现缺损时,对蚁群寻找最佳路径的影响.采用蚁群优化算法进行模拟计算,构建出让耦合体系重新达到最优状态时所对应的拓扑结构,计算出相应的连接通道.研究发现,利用蚂蚁可搜索出新的路径通道,一方面,蚁群可以搜索出替代的路径通道,构建出子空间来弥补局部缺损带来的影响,恢复体系的相应功能;另一方面,蚂蚁搜寻的最佳路径与出现局部毁损的时间长短有关,分别对应于局部最优解和组合式最优解.研究表明,利用蚁群优化算法,不仅可以构建新的子空间修复信息通路,而且可以通过对信息过载现象来预测和排查神经网络或信息网络中的局部病变或故障所在.