One bit of advice is to keep going - you've got a great start. You've done a nice job with the overall description now you need to increase the level of detail. Design is an interative or evolutionary process. You describe what you need, resources available, etc with increasing levels of detail as you go along.
Clearly you'll need to better describe the charging cycle and discharging cycle. Will you simulate or create the ideal charger (optimal charging is quite involved) or try to duplicate current charging methods. Discharging requires some thinking too - will you try to match a pattern that's typical of normal use. Once you've established these things you'll know enough to start thinking about the power supply and the load (for discharging).
It would seem that a lot can be done here. If I were doing this as part of my job I'd map out what I know then consider employing techniques that I learned in design of experiments so that you get the most information out of your work. It sounds like the jig would be computer controlled - PIC, Basic Stamp or maybe a PC or industrial controller depending on complexity.