Travel demand prediction method and system based on hierarchical pattern recognition

The invention relates to the field of travel demand prediction, and discloses a travel demand prediction method based on hierarchical pattern recognition, which comprises the following steps: based onhistorical travel demand data of a scenic area, performing holiday pattern recognition on the travel demand data to search historical time points of the same holiday pattern; searching k modes most similar to the current travel demand mode by adopting different mode time lengths at historical time points with the same holiday mode, and predicting a predicted value of the next point of the currentmode according to predicted values of the next points of the k modes; integrating the predicted values of different mode time lengths, setting the weights of the predicted values under different modetime lengths according to the distance between the current mode and the predicted point, and performing integration to obtain a final predicted value. According to the invention, holiday mode and travel demand mode recognition