|
A Novel Steering Control for Real Autonomous Vehicles via PI Adaptive ( @2 u) Q5 R4 }/ _5 E: l% Z3 E
Dynamic Programming
' A! Z0 F B* T4 x$ m$ G5 \7 ~5 g5 P/ O: F2 u7 z( W& D
$ X' U. N& N' L
! @' F( }* x3 o3 k2 g5 ]" N$ f7 b: ?' R
6 x- {, H3 M( O& q, V8 YA novel steering controller for autonomous vehicles is proposed in this paper. For autonomous vehicles, the efficiency 9 k, |; a' _3 k( Q# M2 c
is an important indicator, in order to reduce the training cost and improve the real-time performance while the car is running on $ |. d1 x) ~3 s* F. E
the road, Offline plus online training is proposed in this paper. Offline plus online training is presented to build a model network 7 a( p5 e) l! S, ^
of vehicle steering and correct it in real time. Then we adopt a policy iteration adaptive dynamic programming (ADP) to get the% S* y7 F/ T s
optimal control law and deploy it to autonomous vehicles. Eventually, the performance of the controller was evaluated using 8 f; w% {5 W: u% H$ f/ p# y
Matlab simulation. Results verify the feasibility of the theory and show the steering controller is one kind of effective, smooth
7 a+ O' Q- @: x; Mcontroller for autonomous vehicles' system.
, V0 O r& |2 n0 t( f$ o% z- \0 T% s+ }/ o; L
/ ]8 ?+ P* O, d/ X) x |