Ion solution to dual random effect

I have been researching  this effect and each time something similar is explained, they only deal with one parameter. My idea has two parameters.

I am looking to make a fire type effect with multicolored LED fixtures. I need to take 5 units and randomly have each at differing percentages of intensity as well as having each a random color from yellow to deep red.

The Intensity levels can be picked before (50, 75,100) and so can the representative colors in palettes (red, orange-red, amber). But the rate at which the colors will transition is slower than that of the intensity changes. Not a flickering flame.

Any suggestions as to how to tie these two randomized effects together?

Related