Anomaly detection algorithm of industrial Internet of Things data platform based on swarm intelligence

The invention discloses an anomaly detection algorithm of an industrial internet of things data platform based on swarm intelligence. The anomaly detection algorithm comprises the following steps of S1, performing preprocessing operation on data acquired by the industrial internet of things IOT data platform; s2, respectively representing different types of collected data; s3, screening abnormal data subsets by utilizing a swarm intelligence algorithm; s4, inputting the obtained abnormal data subset into an anomaly detection algorithm for further training; s5, through multiple rounds of repeated training verification, obtaining different detection results of multiple target objects, and realizing output of data platform anomaly detection; the network is based on a swarm intelligence algorithm framework, the area where important feature data are located can be conveniently positioned in advance, unnecessary calculation is reduced, and the operation speed and performance are improved; meanwhile, a deep learning