<p>The Internet of Vehicles (IOV) represents the integration of the Internet of Things within the realm of smart transportation, serving as a crucial component of the intelligent transportation system. With the emergence of more and more mobile applications for smart cars, a large amount of information collection and computing resources are required. Task offloading, as an effective solution, provides low latency and sufficient computing resources for mobile users in the network. However, how to reasonably offload to reduce system overhead is a challenging issue today. Inspired by parking vehicles on urban roads, this paper establishes a system framework composed of parking clusters, mobile vehicles, edge servers and cloud servers. Considering the system cost delay and energy consumption of task offloading, a weighting factor is set to balance the two parameters, with the goal of minimizing system costs. Request offloading and resource scheduling are analyzed as a dual decision problem. A novel offloading approach for computing tasks in IOV, based on the particle swarm optimization strategy (PSOS), is proposed. The cooling process of the simulated annealing algorithm is employed to probabilistically escape local optima, enhancing the algorithm’s global search capability. Additionally, the evolutionary process of NSGA-II is used to address multi-objective optimization problems. The simulation results show that compared with traditional particle swarm optimization offloading schemes, the proposed scheme in this paper requires lower costs.</p>