A real-time, efficient, and accurate grasp of road network traffic status is the premise of effective control of road traffic networks. Based on the data of the license plate recognition system at road intersections, this paper measures the road network operation status using the congestion index based on the average shortest path length and then uses the K-means clustering algorithm to classify the remaining supply of the network in different periods according to the spectral distance index and the congestion index. On this basis, according to the real-time road traffic data, the congestion index is calculated, and the real-time remaining supply capacity classification of the road network can be obtained. According to the actual remaining supply capacity of the classified real-time road network, different network traffic control schemes are designed, which can effectively alleviate traffic congestion. Taking the local road network of Changsha Wuyi Square as an example, the research results show that the local road network of Changsha Wuyi Square can be divided into four types throughout the day, and the congestion indicators are 1.2,1.3,2.207, and 3.069, respectively corresponding to good, average, poor and extremely poor remaining supply capacity of the road network. In the morning and evening rush hours, the remaining supply capacity of the network is mostly in a poor state, and the network may be in an extremely poor state in the evening rush hours. Compared with weekdays, the remaining supply capacity of the road network performs well in the rush hours on weekends. After the targeted traffic control measures are taken, the simulation results show that the total social cost of congested paths is reduced by more than 5%.