Urban road traffic operation situation typically exhibits fine-grained, strongly time-varying, and multi-period characteristics, traditional traffic quantitative indicators cannot dynamically map the development trend of road traffic operations. To solve the above issue, the traffic tendency evolution index (TTEI) and adaptive KM-CFSFDP cluster algorithm for urban grade roads were proposed. By analyzing the coupling relationship between traffic state and traffic trend and by introducing the dual reward and penalty incentive to get the traffic tendency evolution index. Furthermore, combined with the K-means and clustering by fast search and find of density peaks (CFSFDP) algorithms to model adaptive KM-CFSFDP cluster algorithm, defining three trend types. The experiment on traffic trend of different level of roads in Urumqi for one week indicates that when the cycle ratio of average travel speed is greater and the TTEI is smaller, the road traffic state tends to intensify development. The traffic trend for each road performs the aggravating and stable evolutions in the morning and evening peaks on work days and non-work days. Additionally, the validity and accuracy of adaptive KM-CFSFDP cluster algorithm was verified compared with three baselines. This research can offer valuable insights for the overall perception of urban road traffic operation situation.