You are able to read a file and write to a new file? So that is the basic part of your program.
Then you need to decide how to identify, from all the records you have read, which ones you want to process.
So, think how you would find the IDENTIFICATION DIVISION. As Robert pointed out, remember that you don't have to spell IDENTIFICATION, you can shorten it to ID (I always did, it makes the program look so much neater).
Generally there is a place you would expect to find this starting (I'm not doing the whole job for you). When you identify it, follow enrico's advice.
You have now provided a little more information on your copybook or dclgen. When you identify them, again follow enrico's advice.
So, how to find them. You have mentioned UNSTRING. OK, you should be able to get it to work. Look it up in the manual. Try to understand it. Try some examples. Make sure you know how it is handling "leading space" - the space at the start of each line. Then transfer the code idea to your reading/writing program.
It'll be a very, very, basic thing. Is this a training task? Or does it have some deeper purpose?