Deep learning improved algorithm for cold source disaster prediction

Abstract

The invention discloses a deep learning improved algorithm for cold source disaster prediction, and the algorithm comprises the steps: carrying out the statistical classification of related factors affecting cold source disaster, carrying out the normalization of multi-source heterogeneous data through a data normalization function, and obtaining grouped sample data; establishing a deep belief network, adopting a deep belief network improved algorithm based on the momentum learning rate, carrying out the repeated iterative training of model parameters in combination with the multiple sets of sample data, and acquiring a determined mapping relation between input and output of the deep belief network; and pre-estimating the disaster level of the cold source disasters. According to the method, the improved algorithm of deep learning is adopted, so that the network learning efficiency is effectively improved, the error convergence rate of prediction is reduced, and the problem of prediction of ocean disasters with nonlinearity, time-varying property and uncertainty is well solved.

Description

Technical field[0001]The invention belongs to the technical field of marine disaster early warning, and specifically relates to an improved deep learning algorithm for cold source disaster prediction.Background technique[0002]In recent years, with changes in the marine environment, a large amount of garbage, jellyfish, fish, seaweed, seaweed, etc. have arrived at the cold source water intake of the coastal nuclear power plant along with tides and wind waves. In order to reduce the threats caused by foreign matter in the ocean to the safe operation of coastal nuclear power plants, disaster prediction in the sea area around the water intake of the coastal nuclear power plants has become an urgent issue for the safety of nuclear power plants’ cold sources.Summary of the invention[0003]The purpose of the present invention is to provide an improved deep learning algorithm for the prediction of disasters caused by cold sources, which can accurately predict the level of disasters that may ...

AI Technical Summary

Problems solved by technology

[0002] In recent years, with the change of the marine environment, a large amount of garbage, jellyfish, fish, seaweed, seaweed, etc. have been sent to the cold source water intake of the coastal nuclear power plant with the tide, wind and waves, which will cause the nuclear power unit to reduce power or even stop the reactor in severe cases.

Images