Passing current trajectory step to select_action function

[Federico-PizarroBejarano]

In some controllers we are getting the current step of the trajectory from the environment, via the info, such as

safe-control-gym/safe_control_gym/controllers/pid/pid.py

Line 94 in 0d91ee2

step = self.extract_step(info)

In others, we keep track of the step in the controller itself via self.traj_step, such as

safe-control-gym/safe_control_gym/controllers/mpc/mpc_acados.py

Line 337 in 0d91ee2

self.traj_step += 1

We should choose one and standardize it.

The benefit of the info approach is that the iteration is coming from the environment, and thus the controller is stationary, i.e., if you call a select_action function twice with the same arguments you will get the same action. The drawback is that in real experiments on real hardware, there is no info dictionary since the environment isn't giving you more information. One way around this is creating an info dictionary in the real experiment script and passing that in with the relevant data.