<div class="section abstract"><div class="htmlview paragraph">Internet of vehicles (IoV) system as a typical application scenario of smart
city, trajectory planning is one of the key technologies of the system. However,
there are some unstructured spaces such as road shoulders and slopes pose
challenges for trajectory planning of connected-automated vehicle (CAV).
Therefore, this paper addresses the problem of CAV trajectory planning affected
by unstructured space. Firstly, based on cyber-physical system (CPS), the
cyber-physical trajectory planning system (CPTPS) framework was built. A
high-precision digital twin CAV is established based on the physical properties
and geometric constraints of CAV, and the digital model is mapped to cyber space
of the CPTPS. In order to further reduce the energy consumption of the CAV
during driving and the time spent from the start to the end, a model was
established. Further, based on the sand cat swarm hybrid particle swarm
optimization algorithm (SCSHPSO), global path planning for connected-automated
vehicles is performed; The vehicle trajectory is smoothed based on a Bezier
curve. Finally, the simulation results show the trajectory planning results in
unstructured space and two-dimensional plane. Compared to the sand cat swarm
optimization (SCSO) algorithm, the fitness function value of the trajectory
planned by the SCSHPSO algorithm in unstructured environment has decreased by
6.34%. The simulation results demonstrate the performance of the CPS based
trajectory planning scheme for connected-automated vehicles designed in this
paper, especially in unstructured environments, where the SCSHPSO algorithm is
more competitive.</div></div>