With a bunch of them and a similar number of diodes (assuming your input is AC) you could build a voltage multiplier.
But at 2-3 volts that would be HORRIBLY inefficient, due to diode drop. You'd lose maybe 0.6 volts to the diode, and more to vlotage droop from the capacitor, in every step. That's dumping maybe a third of your power for each volt or two you gain.
Further, if you're thinking about using this on a mill when the wind is low, the FREQUENCY is very low, requiring an ENORMOUS capacitor. A farad is big for mid-to-high audio rates and signal (or even audio power) current levels. But it's miniscule for low-freuqency power applications.
But there's another problem with trying to get power from a windmill at low winds: The power AVAILABLE is proportional to the CUBE of the wind speed. So at low winds you not only have negligible voltage (from slow rotation) but negligible current (from low force on the shaft). Boost circuits trade current for voltage, and with little current BEFORE you start trading it away, you're left with dribbles.
And you spent a lot of effort and equipment to get those dribbles. And maybe changed your system to make it less reliable, or less efficient on TYPICAL winds, when there's lots of power to collect - and thus a lot to be LOST by even a small reduction in efficiency from tuning for lower speeds.
If you need more power your resources are normally much better spent to increase the size of the mill (collecting more power at typical wind speeds) and your storage (to save it for low-wind times) than to try to extract a few drops from the mud when the stream has run dry.