I have 100s of inventions on paper in short hand. I have improvements on the rubber glove to improvements on the electric generator. I do not know how to pick a winner and go with that one first? Should I go with the cheapest to make first?
Consider what, if patented, would really bring you commercial value.
If you're never going to actually make and sell the patented device, then what's the point of applying for a patent? For example, are you really going to make and sell an electric generator?
The exception is if you want to be a non-practising entity. You can then sell your patented improvement to others, or sue others when they independently implement it. The former is very rare (except perhaps in pharmaceuticals and telecommunications). The latter is somewhat more common, but depends on whether you want a livelihood built on shaking down legitimate businesses.
Moreover, consider why you want a patent at all. If you haven't commercialised any of your ideas already, then perhaps a lack of a patent isn't the underlying issue. A business without a patent is better than no business with a patent.