Commercial vehicle driving fuel consumption evaluation method based on scene

The invention relates to the technical field of big data analysis, and discloses a scene-based commercial vehicle driving fuel consumption evaluation method, which comprises the following steps: step S101, constructing a first driving sequence; s102, constructing a second driving sequence based on the first driving sequence; s103, inputting the second driving sequence into the prediction model, wherein an output value represents a probability value belonging to different driving scenes; step S104, constructing a driving scene library, wherein each driving scene corresponds to a minimum fuel consumption value and an average fuel consumption value; step S105, obtaining an average fuel consumption optimization potential value and a maximum fuel consumption optimization potential value of the current vehicle based on the driving scene library; according to the method, information transmission and fusion are carried out on the driving data of the vehicle in the time dimension, the driving distance of the vehicle i