CarND PID Control

Self-Driving Car Engineer Nanodegree Program

---

You can find the Clojure version of Udacity's PID controller project from term 2 of the self-driving car engineer nanodegree here. All thanks to Eric Lavigne: He's a great software architect.

How to find the right parameter for the PID-controler:

I did this process by hand. First search find good values for P and D while I is still zero at 30mph. This is the case at 30mph for example

P = 0.1

D = 2.0

I = 0.0

For a speed of 60mph increase the D value, while lowering the P value.A small I value improves the result a bit.

P = 0.05

D = 4.0

I = 0.001

Finaly at 70mph a optimal value of I is neccesary to drive around the corner.

P = 0.025

D = 5.0

I = 0.002

Speed = 70MPH:

Conclusion

The new starter code in Clojure has a speed of 70mph for the lake trac, while the old starter code in C++ has a speed limet of 50mph only.