|
In order to reduce the negative impact of random time-varying delays and packet losses on cooperative driving for largescale connected and automated vehicles (CAVs) at communication-constrained unsignalized intersections, a cooperative driving method based on improved model-free adaptive predictive control (IMFAPC) is proposed in this paper. The control method designed in this paper only used the input and the output (I/O) data of the CAVs during driving based on pseudopartial derivative (PPD) dynamic linearization technology, which avoids the complex, nonlinear and uncertain modeling process of multi-vehicle system, as well as the system control issues caused by the use of imprecise and uncertain multivehicle system control model. The total time-varying transmission time-delay and packet loss between the target on-board unit (OBU) and road side unit (RSU) through the vehicle to infrastructure (V2I) communication is comprehensively considered. The control objectives can be achieved by the networked predictive control of time-varying transmission timedelay and packet loss, and the expected speed is allocated for each target vehicle through the RSU with edge computing (EC) controller. Compared with the two benchmark methods under the time-varying delay and packet loss conditions, the results show that the proposed method has effectively reduced the influence of random time-varying delay and packet loss on the control effects, reduces the safety risk of multi-vehicle driving and improves the traffic efficiency under various traffic volumes.
|
