Robust optmization of distribution networks with DGs using two different patterns random clustering

Robust algorithm is known as the one of high potential models for strengthening in optimization of large and complex distribution networks considering uncertainty. In this research we are looking for some suitable template and appropriate clustering pattern in order to analyze the operation of distribution network. The analysis of optimization process in both individual and adaptive cases of distributed generation sources applied to some standard distribution network shows that the adaptive case of DGs determines less partial and total operation cost. except in islanded case of DGs. The results of clustering approach based on choosing the candidate points which includes DGs and feeders are used in medium voltage (MV) network level. By investigating the software data in both single cluster with first up to fifth and tenth repeatition, one to five clusters and finally ten clusters case shows the superiority of mean data besed clustering. because of having the three following properties such as being sensitive to all weighted candidate points, the speed of acceptable operation and feeding the clusters in islanded case in multiplicity of clusters, it is suitable to designate the robust optimization method using pattern random clustering.