Aiming at the problem of task offloading failure caused by the excessively fast vehicle moving in the Internet of vehicles, an effective task offloading risk assessment model was designed, and a dynamic task offloading scheme based on joint resource allocation was proposed. Time, energy consumption, and risk were jointly modeled as system utility, and the system utility was maximized through joint optimization of offloading decision and resource allocation. The optimization problem was formulated as a mixed integer nonlinear programming. Convex optimization techniques were used to solve the computational resource allocation problem under a given offloading decision, power allocation was optimized through fractional programming technology. The simulation analyzes the impact of vehicle mobility on system utility, and the rationality of the proposed scheme is proved.