In order to analyse the trend of crude oil prices more effectively, this paper takes Daqing crude oil daily spot price time series as the research object, combines the econometric model with the complex network method, defines the autoregressive mode quantitatively, uses the sliding window idea to set the reasonable form length and the step number, divides the time series into several sub-patterns, and establishes several autoregressive sub-patterns. The autoregressive sub-patterns is set as a node, the transmission between the sub-patterns is set to the edge, the autoregressive sub-pattern transmission complex network is established, and the characteristics and properties of the complex network are used to study the spot price time series characteristics of Daqing crude oil. We find that a small number of autoregressive sub-patterns and transmission modes drive the oscillation of spot price time series of Daqing crude oil, and the clustering effect on fluctuation appears in the transmission process, and some non-main autoregressive sub-models have high intermediary ability in crude oil time series. This work not only proposes a new analysis for the crude oil prices time series but also provides important information for investors.