.png){kind=icon}
.png){kind=icon}
Return
Variable step MPC trajectory tracking control method for intelligent vehicle
.png){kind=icon}
PRE
AI
DOI：10.1007/s11071-024-10042-x.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
To improve the accuracy, real-time and stability of intelligent vehicle path tracking control algorithms, a variable Step Model Predictive Control method (VMPC) for path tracking based on Model Predictive Method (MPC) is proposed. A vehicle dynamics model considering path tracking was constructed, and a VMPC controller was designed based on the model. To address cumulative model error, the proposed control method employs a zero-order holder-based short-step discretization prediction model in the front part of the prediction interval and a first-order holder-based long-step discretization prediction model in the back part. Carsim/Simulink co-simulations were conducted to compare the performance of the proposed VMPC controller with that of a traditional MPC controller on double-lane roads and highways. The simulation results indicate that the proposed VMPC controller exhibits superior control precision, smoothness, real-time performance, and dynamic stability. The proposed method decreases 56.6% for the lateral error, 52.4% for the heading error, 28.5% for the sideslip angle, and 45.7% for the average solution time at most when compared to a standard MPC. Experiments were performed on a drive-by-wire integrated chassis platform, which confirmed that the proposed VMPC controller achieves desired tracking control accuracy for variable curvature paths in engineering applications.
Keywords:
Motion control
Path tracking
Variable step model predictive control
Model predictive control
Journal
IF:
6
Papers:
1.4W
Citations:
4.1W