Safe-RL algorithm implementation

[rickstaa]

We need to implement the Safe-Rl algorithm around the RL-algorithm that was implemented in the environment.

[rickstaa]

These save learning algorithms can be found in the bayesian_learning_control package.