Investigation of Stored Energy Distribution in Filters Using K-Means Clustering Algorithm

The k-means clustering algorithm has been implemented to find patterns in the time-averaged stored energy distribution in various filter networks. A large data set comprising of numerous topologies for 50 different single band specifications has been investigated. By finding key characteristics within this data set, general guidelines for predicting the optimum topology for power handling have been established.