Cooperative sensing system based on multilayer distributed learning and continuous learning method

The invention discloses a collaborative sensing system based on multilayer distributed learning and a continuous learning method. A cloud center of the system comprises a cloud center server and an edge aggregation server. The cloud center server is used for carrying out dynamic clustering on each edge node to obtain clustering grouping information and sending the clustering grouping information to the edge aggregation server; the edge aggregation server is used for selecting an edge active learning platform corresponding to each edge node according to the clustering grouping information and adding the edge active learning platform into an edge cluster so as to obtain an edge aggregation model; and the edge active learning platform is used for receiving the edge aggregation model and loading the edge aggregation model to the unmanned terminal platform corresponding to each edge node. According to the method, the edge nodes are dynamically clustered through the cloud center, so that aggregation is completed on