Cyanobacterial bloom prediction method based on recursion time sequence deep belief network

Abstract

The invention discloses a cyanobacterial bloom prediction method based on a recursion time sequence deep belief network, and belongs to the technical field of water environment prediction. Firstly, water quality data are collected, preprocessed and divided into training data and test data, a multi-factor input and single-factor output RTDBN model is constructed, and a hidden layer 2 is establishedon the RTDBN model to perform feature extraction on the training data again. For the training data, a parameter relationship between an input layer and a hidden layer 1 is established on the RTDBN model, and the input layer bias and hidden layer 1 bias are updated by adopting CRBM; and establishing a parameter relationship between the hidden layer 1 and the hidden layer 2. And then the new biasof the hidden layer 1 and the bias of the hidden layer 2 is updated, further the characteristics of the hidden layer 1 by the hidden layer 2 is extracted, and the weight of the RCRBN is updated, and finally, the fine-tuning of model parameters is reversely carried out, RTDBN model training is completed, and cyanobacterial bloom of water quality is predicted through test data. According to the invention, the calculation of the model is reduced, the over-fitting phenomenon is prevented, and the prediction precision is improved.

Description

technical field [0001] The invention belongs to the technical field of water environment prediction, and relates to a water bloom prediction method, in particular to a cyanobacteria bloom prediction method based on a recursive time series deep belief network. Background technique [0002] Water quality characterizes the physical, chemical and biological characteristics of water and can be used to measure the usability of water bodies to society. Eutrophication refers to the process of excessive enrichment of nutrients in water bodies, resulting in accelerated reproduction of certain biological productivity. Symptoms of eutrophication mainly include algal blooms and eutrophication of water bodies. Eutrophication of water body is a natural process, but human activities can accelerate this process by increasing the load of nutrients entering the water body. Eutrophication enrichment will lead to algal blooms, which will destroy the process of aquatic ecological balance. There...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the existing problems such as low precision of water bloom prediction and insufficient number of samples, the present invention combines the improved deep belief network method with a time series model to construct a cyanobacteria bloom prediction based on a recursive time series deep belief network method, thereby improving the prediction accuracy of algal blooms, and providing a new idea for the prediction of algae blooms in lakes and reservoirs

Images