Properly tuned the new PID parameters

[ryan-summers]

Parameters were tuned using the Ziegler-Nichols method. I'm really not sure how the old parameters happened to pass the tests with improper derivative calculations, but whatevs! This should fix the heartache of the other AI devs :P The simulator_bridge frequency was increased to ensure that the IMU node publishes faster than 5 Hz (this causes a large lag in the derivative calculation and causes the control system to perform sub-optimally).

---

This change is 

[irwineffect]

Reviewed 2 of 2 files at r1.
Review status: all files reviewed at latest revision, 1 unresolved discussion, some commit checks failed.

---

param/simulator.yaml, line 8 at r1 (raw file):

        #  such as control, depth, sensor in cobalt_sim.yaml and
        #  any rate parameters below
        simulator_bridge: 150

Why is this so high?

---

Comments from Reviewable

[skallaher]

@irwineffect, could you complete your review here so that this can be merged in? @rsummers, the roll test of the control system is currently failing on Travis, is there an explanation for this? Does the test need to be updated?

---

Review status: all files reviewed at latest revision, 1 unresolved discussion, some commit checks failed.

---

Comments from Reviewable