|
A Novel Steering Control for Real Autonomous Vehicles via PI Adaptive * C- x' l2 [& H( j4 `4 q
Dynamic Programming
! L- W5 a. F3 f w! D1 b
# U( C p' X* k# R5 o+ L* c3 b7 H# v6 T8 |$ C u' F+ V
2 t. A5 d: _2 d) ^
8 _ k$ t( a4 d$ ]4 q. N
2 v0 i# H" m) b+ X2 @ tA novel steering controller for autonomous vehicles is proposed in this paper. For autonomous vehicles, the efficiency - b7 @; B. n, J
is an important indicator, in order to reduce the training cost and improve the real-time performance while the car is running on
( C- E& k: Y( `5 Gthe road, Offline plus online training is proposed in this paper. Offline plus online training is presented to build a model network
2 B( B1 u3 r0 H R* |of vehicle steering and correct it in real time. Then we adopt a policy iteration adaptive dynamic programming (ADP) to get the
* J7 |: d2 `' i& [, Soptimal control law and deploy it to autonomous vehicles. Eventually, the performance of the controller was evaluated using ) r" z, l) D8 {( f' h# S) C
Matlab simulation. Results verify the feasibility of the theory and show the steering controller is one kind of effective, smooth
2 K) ]; J: R9 w5 b( O, Z$ Kcontroller for autonomous vehicles' system. 6 R3 d0 m( }! b6 i- w3 O% g
9 H. T0 ~7 B/ P+ A1 p% {( G! B$ g3 ?( z4 z4 S- f% G
|