Such plans can be tricky for mathematical work as it does not always follow a predictable path or timescale, as much as we would like it to. I'm interested in what others say but this is my view/approach (applied math/engineering):
First, take a step back and consider what the purpose of this plan is. The plan should give you (and anyone who's reading it) confidence that what you want to do is achievable within the time available. It should also provide a way of regularly assessing progress towards meeting the objectives.
With this in mind, breaking the project into steps will be very problem specific but goes something like this: Imagine the project runs perfectly (!), identify the techniques/theorems/etc you intend to use to solve the problem. How will you adapt/use/apply these to your problem? What are the challenges to overcome for this to work and how will you overcome them? This should give you some steps to discuss as well as some indication of what the intermediate and final outputs will be (a proof, numerical model, etc).
It's rare for a project to run perfectly. So now that you have a perfect plan, it's time to introduce contingencies for things not going as expected. Give yourself much more time than you think you will need for each task (but still realistic). When describing each task, discuss what you will do if your first attempt does not work. What will you try next? Is there a 'last resort' that is 'guaranteed' to work? Perhaps a computer simulation or making a simplifying approximation which has been used previously.
If you don't know what approach you will use, you may not be in a position to write such a plan yet. You may need to do some initial calculations/research to formulate a realistic plan.