|
A Novel Steering Control for Real Autonomous Vehicles via PI Adaptive
! W6 D3 V. u. l; e9 kDynamic Programming , s4 m# b: I. H/ r, T$ Y' d L
: c& ^- M/ A% U7 n# G: B
6 p5 p3 O, c' p* _4 r4 k
$ V, j* w' ~" W- {! _8 ^& a* A9 @( f# q2 W! N; y
; C8 Z; B, i% a; B' t% V/ B* A( Z# wA novel steering controller for autonomous vehicles is proposed in this paper. For autonomous vehicles, the efficiency $ C: U; z* I# i! f* `- Q
is an important indicator, in order to reduce the training cost and improve the real-time performance while the car is running on 8 w5 X3 w1 `, N4 U$ @6 s( P# X' L
the road, Offline plus online training is proposed in this paper. Offline plus online training is presented to build a model network 1 S- H& ~4 F1 j: x& z0 D
of vehicle steering and correct it in real time. Then we adopt a policy iteration adaptive dynamic programming (ADP) to get the
5 m- Z3 `. x* f; S+ Aoptimal control law and deploy it to autonomous vehicles. Eventually, the performance of the controller was evaluated using
8 ?# V+ w o/ x9 fMatlab simulation. Results verify the feasibility of the theory and show the steering controller is one kind of effective, smooth + N2 b, f& s ^' R; b0 h7 ~; Z7 S
controller for autonomous vehicles' system. . Z9 F6 L# p. [) T( x& l$ }
" y% J2 {( A) o& \/ P, r6 h- R* U
3 S4 M* ?4 |& g( I |