The algorithm isn't awfully complex. If it's 6pm, and you need your car charged by 6am, and you need to be charging for 6 hours, then you just purchase the cheapest 6 energy futures for the 6 hours you know you'll need, and don't purchase the other 6.
If at any time during the night the prices of the unpurchased futures drop below that of the ones you have purchased, you can sell one and buy a different one (rescheduling the charge).
There isn't any better algorithm within those constraints. As long as the energy futures market is a fully liquid market, the price you'll get will be (on average) equal to the spot price you would have paid if you'd made the ideal scheduling decisions and bought on the spot market.
Fair point. For car charging the algorithm can be relatively simple (if suboptimal due to imperfect assumptions like "fully liquid"), but for stationary batteries + solar it gets much more complex.
I still suspect more sophisticated players will find easy ways to "game" such an algorithm (eg pre-bidding). Whenever you have a large herd of predictable sheep, it's an opportunity for a wolf.
If at any time during the night the prices of the unpurchased futures drop below that of the ones you have purchased, you can sell one and buy a different one (rescheduling the charge).
There isn't any better algorithm within those constraints. As long as the energy futures market is a fully liquid market, the price you'll get will be (on average) equal to the spot price you would have paid if you'd made the ideal scheduling decisions and bought on the spot market.