[Enhancement]: Migrate normalization of the action_space to a wrapper.

[AlejandroCN7]

Improvement 🔧

By default, normalization in action space is activated in our environments due to the fact that is required for some algorithms directly. However, I think it is a more general solution to do this work in a wrapper and make native Sinergym environments as simple as possible.

Original behavior

There is an attribute flag in Sinergym environments (default to True), to normalize observation and parse it to original values before to send the values to the simulator

Enhanced behavior

Encapsulate all this functionality into a wrapper, and write documentation to warn that this wrapper is required/recommended for some DRL algorithms.

Additional context

Add any other context about the problem here.

Checklist

• I have checked that there is no similar issue in the repo (required)
• I have read the documentation (required)

📝 Please, don't forget to include more labels besides enhancement if it is necessary.