City anomaly detection method and device based on probability distribution

The invention discloses a probability distribution-based city anomaly detection method and device. The method comprises the following steps of: dividing a city region according to city road network data to obtain city sub-regions; vehicle track data are obtained through preprocessing; according to the vehicle trajectory data, time period traffic inflow and time period traffic outflow of the urban sub-regions in historical 30 days are counted, and then an inflow probability distribution set and an outflow probability distribution set of the urban sub-regions are constructed by using an FPD method; detecting and marking anomalies in the incoming flow probability distribution set and the outgoing flow probability distribution set by using an LOF algorithm; obtaining a current detection database of the to-be-detected city sub-region by finding out the approximate region; according to the current detection database, using an LOF algorithm to determine whether there is an abnormal event in the to-be-detected city s