search

svineet / pid-simulation

PID controller simulation and subsequent control using Reinforcement Learning algorithms.
3 stars 0 forks source link

Reward function ideas #5

Open svineet opened 4 years ago

svineet commented 4 years ago

Idea 1

V = k1*(norm of error)^2 + k2 *(norm of change in error)^2
exp(- V/sigma)

Idea 2

Abhinava's idea from WhatsApp group

svineet commented 4 years ago

Requirements for reward function: