The invention discloses a bus passenger flow evolution analysis method based on time-space data fusion. The method comprises the following steps: 1, a historical time-space data preprocessing step; 2, a step of calculating trip get-off probabilities; 3, a bus passenger flow time-space evolution analysis step; and 4, a multi-dimensional passenger flow time-space distribution feature analysis step, i.e., according to an obtained time-space network and flow of corresponding time-space network arc segments, drafting a bus time-space distribution feature graph. The analysis method provided by the invention has the following advantages: 1, the trip get-off probabilities are obtained based on enormous historical time-space data statistics, the precision is quite high, the scheme is simple and practical, the solving efficiency is high, and correlation operation indexes of a whole city bus system can be rapidly solved; and 2, the real-time performance is good. Besides, based on the calculated and stored trip get-off probabilities, station OD is calculated in real time through combination with bus get-on passenger flow, and accordingly, the time-space evolution rule of line passenger flow can be dynamically reflected in real time.