Cyanobacterial bloom prediction method based on multi-factor time sequence-random depth confidence network model

Abstract

The invention discloses a cyanobacterial bloom prediction method based on a multi-factor time sequence-random depth confidence network model and belongs to the water environment prediction technologyfield. In the invention, improved depth confidence network method is combined with a multi-factor time sequence analysis method so as to construct the multi-factor time sequence-random depth confidence network model, and then, a MT-RDBN model is adopted to discretize, an RCRBM learning algorithm and MT-RDBN model parameter fine-tuning are adopted too. An autoregressive model and a multi-factor regression model in a time sequence are used when the model is established, influence factors are considered so that an MT-DBN model can predict the representation factor of a future moment through the chlorophyll concentrations and the influence factor data of a current time and a historical moments, a sample application amount is reduced and prediction precision is increased.

Description

technical field [0001] The invention relates to a water bloom prediction method, which belongs to the technical field of water environment prediction. Specifically, it is a cyanobacteria bloom prediction method that improves the prediction accuracy of a multi-factor time series-random deep belief network model established after analyzing the generation process of water blooms. Background technique [0002] With the development of society and economy, the phenomenon of eutrophication in water body is becoming more and more common, seriously affecting people's normal life. The eutrophication phenomenon of water body is a natural phenomenon that occurs in fresh water and is caused by the sudden excessive proliferation of algae due to the high content of nitrogen, phosphorus, and potassium in the water body. The eutrophication of water bodies is mainly caused by the discharge of nitrogen, phosphorus, potassium and other elements into surface water bodies with slow flow rate and...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the problem that the existing water bloom prediction accuracy is not high, and the prediction problem of the highly nonlinear system cannot be processed only by a single factor or a single mathematical model method, and the improved deep belief network method and multi-factor time series analysis A multi-factor time series-random deep belief network model is constructed by combining the method and method, which can improve the prediction accuracy of algae blooms and provide a new idea for the prediction of cyanobacteria blooms

Images