I've read that the most of the problems can be solved with 1-2 hidden layers.
How do you know you need more than 2? For what kind of problems you would need them (give me an example)?
Artificial Intelligence Stack Exchange is a question and answer site for people interested in conceptual questions about life and challenges in a world where "cognitive" functions can be mimicked in purely digital environment. It only takes a minute to sign up.Sign up to join this community
Formally, a single hidden layer is sufficient to approximate a continuous function to any desired degree of accuracy, so in that sense, you never need more than 1. This is called the Universal Approximation Theorem.
Finding the best topology for a given problem is an open research problem. As far as I know, there are few universal 'rules of thumb' for this.
For a given problem, one option is to apply a neuroevolutionary approach such as NEAT, which attempts to find a topology that works well for the problem at hand.