TY - JOUR
T1 - Reconfigurable Model Predictive Control for Articulated Vehicle Stability with Experimental Validation
AU - Zhang, Yubiao
AU - Khajepour, Amir
AU - Hashemi, Ehsan
AU - Qin, Yechen
AU - Huang, Yanjun
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/3
Y1 - 2020/3
N2 - This article proposed a reconfigurable control scheme for articulated vehicles' stabilization by leveraging optimization-based control techniques. The central objective is to maintain a good lateral and yaw stability of the vehicle with optimal corrective brakes, meanwhile applicable to different actuation configurations. This is achieved by a two-layer control structure, where the high-level controller formulates as a model predictive control (MPC) tracking problem to generate corrective center-of-gravity (CG) yaw moment of each unit. The lower level controller utilizes the control allocation (CA) algorithm with real-time constraints to optimally calculate differential brakes at each wheel with maximum utilization-of-tires capacity. To evaluate its real-time performance, experimental validation is carried out on the electrified tractor-trailer with selective differential braking systems. It is observed that the controller is effective in dynamics control, meanwhile reconfigurable to various actuation configurations. Furthermore, the proposed system has great potential in production tractor-trailer systems due to the low cost and number of sensor requisites.
AB - This article proposed a reconfigurable control scheme for articulated vehicles' stabilization by leveraging optimization-based control techniques. The central objective is to maintain a good lateral and yaw stability of the vehicle with optimal corrective brakes, meanwhile applicable to different actuation configurations. This is achieved by a two-layer control structure, where the high-level controller formulates as a model predictive control (MPC) tracking problem to generate corrective center-of-gravity (CG) yaw moment of each unit. The lower level controller utilizes the control allocation (CA) algorithm with real-time constraints to optimally calculate differential brakes at each wheel with maximum utilization-of-tires capacity. To evaluate its real-time performance, experimental validation is carried out on the electrified tractor-trailer with selective differential braking systems. It is observed that the controller is effective in dynamics control, meanwhile reconfigurable to various actuation configurations. Furthermore, the proposed system has great potential in production tractor-trailer systems due to the low cost and number of sensor requisites.
KW - Control allocation (CA)
KW - electrified articulated vehicles
KW - model predictive control (MPC)
KW - real-time experiment
KW - reconfigurable control
KW - transportation safety
UR - http://www.scopus.com/inward/record.url?scp=85082512530&partnerID=8YFLogxK
U2 - 10.1109/TTE.2020.2972374
DO - 10.1109/TTE.2020.2972374
M3 - Article
AN - SCOPUS:85082512530
SN - 2332-7782
VL - 6
SP - 308
EP - 317
JO - IEEE Transactions on Transportation Electrification
JF - IEEE Transactions on Transportation Electrification
IS - 1
M1 - 8986684
ER -