In the process of realization of vehicle-road cooperation technology, intelligent vehicle penetration and replacement and comprehensive layout of intelligent roadside facilities need a long period of time, and it will be in a state of incomplete vehicle-road cooperation for a long time in the future. In this state, the edge computing server has limited computing power,and tasks such as signal-control, special service priority, and unmanned driving in the road network run concurrently,which need a system providind low latency of less than 100 milliseconds.However,the existing coordination methods don’t consider the criticality and timeliness of different tasks, causing important tasks to be queued without allocating enough resources. Based on the above problems, this paper proposes a intersection control task scheduling model based on the criticality and timeliness.Vehicles in incomplete vehicle-road cooperation are divided into four categories: Intelligent vehicle, Ordinary car, Bus and Emergency Vehicle.Each type of vehicles carries different types of control task.We defines the task of criticality and timeliness.Besides, build a time window,and established a vehicle-arrival model, a communication model and a calculation model between vehicles and the edge computing servers. We make use of traffic volume and network environment to carry out dynamic scheduling to ensure priority execution of tasks with high criticality and timeliness, and obtain the corresponding minimum system delay. KEYWORDS: Edge Computing ; vehicle-arrival model; dynamic Task Scheduling; criticality and timeliness;

CICTP