There are methods like fuzzy c-means, subtractive clustering, grid partitioning, etc. But the answer depends on whether you can clearly identify the centers for the distribution of data or not. Could you provide a little more information about the data distribution so that other fuzzy logic experts can help?