The invention provides a method and
system for quickly identifying
power consumption abnormal behaviors. The
system comprises the following modules: a
power consumption difference
trend analysis module which calculates
power consumption difference trends based on a power consumption information
big data analysis method, compares the similarity of the power consumption difference trends between different users and different periods, and deduces the probability of power consumption anomaly; a power utilization curve
similarity analysis module which is used for establishing a power consumption quantity curve and a
line loss curve of lines, users and
transformer areas, calculating the form similarity of the power consumption quantity curve and the
line loss curve, and calculating the similarity of various curves by using a
shortest distance method; and a
transformer area
user information anomaly analysis module which is used for analyzing a
transformer area power supply quantity, a power selling quantity and transformer area
line loss by utilizing a
big data technology, and calculating the probability of transformer area
user information anomaly on the basis of parameters such as the transformer area maximum power supply
radius and the transformer area
linear distance. With the method and
system adopted, the quick recognition of the abnormal power consumption behaviors can be realized, the abnormal probability of the
user information in the transformer areas can be calculated, and the
troubleshooting workload is reduced as much as possible, and a power failure range can be narrowed as much as possible.