Which neural network architecture to choose for turning control?

Question

Neural networks

The bottom line is this: we need an adaptive system based on a neural model, because having figured out the issue, modern adaptive CNC control systems are not quite "adaptive", they are still based on empirical algorithms and are not very accurate. The idea is that there are diagnostic signs in the cutting process, according to the analysis, I came to the choice of such a parameter as cutting force, it is very easy to measure and a rather informative parameter (of course there are some disadvantages, but this is the easiest option). Therefore, the cutting force will be applied to the input, and the output needs:
1. Cutting speed
2. Feed
And possibly 3. Depth of cut (but in theory it is possible without it)
I am not very competent in the matter of neural networks, so I can write further nonsense, but I really want to understand this topic.
So, the neural network must have memory and feedback (that is, it turns out a recurrent neural network), I read that they are mainly used for language tasks, so I don’t even know where to read on the topic of recurrent ones for process control. I hope this time I painted in more detail, if I add anything else