Clustering of time series data

I have a time series data set. I want to use Dynamic time warping for distance measurement. For algorithm, I was thinking of using either K-means DTW Barycenter Averaging (DBA) or K-medoids. Data has outliers. My goal is to identify demand pattern. I am not sure which one to use. Which one would be better in terms of accuracy and evaluation? What are the advantages and disadvantages of both of the algorithms and what type of validation should I use? I would like to have a visual representation of the outcome, not statistical validation.