|
A Novel Steering Control for Real Autonomous Vehicles via PI Adaptive ! u! a! }) U3 N9 o
Dynamic Programming # {" {& y" e( a; V# M
+ o2 }9 f8 i) Y: ?8 f; n
- G0 \5 t: o: s
1 T% o- T; ?- m! p7 B
2 z5 Z' L$ O# Q! V2 l/ e4 J* I" }- d1 y
A novel steering controller for autonomous vehicles is proposed in this paper. For autonomous vehicles, the efficiency & |0 m0 G( }7 P5 T, U' g3 Y: u
is an important indicator, in order to reduce the training cost and improve the real-time performance while the car is running on - K! S& t. v' W$ I' y) X
the road, Offline plus online training is proposed in this paper. Offline plus online training is presented to build a model network 2 m$ ~: i* {2 ~6 ?- O3 i
of vehicle steering and correct it in real time. Then we adopt a policy iteration adaptive dynamic programming (ADP) to get the
6 P0 n4 U# i6 i; qoptimal control law and deploy it to autonomous vehicles. Eventually, the performance of the controller was evaluated using
) Z% k5 g$ V2 l" |& K& kMatlab simulation. Results verify the feasibility of the theory and show the steering controller is one kind of effective, smooth
1 Q( I. m( Q5 ]controller for autonomous vehicles' system.
7 ^9 M0 Y' G# s8 W: d
. g) V ~3 w" }( o8 `
" ^ u7 H8 H4 l6 Z$ V& @! c |