A method and device for predicting ammonia nitrogen in water body based on deep belief network

Abstract

The present invention provides a water body ammonia nitrogen prediction method and device based on a depth belief network. The method includes: based on the actual water quality parameters and actual environmental factor parameters of the water quality of the target water body, using a pre-established depth belief network prediction model to predict the target The ammonia nitrogen content of the water body; wherein, the depth confidence network prediction model is initialized in advance according to the prediction demand, and utilizes the selected sample data to carry out training and testing based on the temporary abandonment point dropout, and the sample data is the sample water quality environment The parameters are aligned with the ammonia nitrogen content of the sample water. The invention is simple and easy to implement, can effectively improve the prediction operation speed and prediction accuracy of the ammonia nitrogen content in the water body, and effectively improve the stability and universality of the prediction method.

Description

Technical field [0001] The present invention relates to the field of computer technology, and more specifically, to a method and device for predicting ammonia nitrogen in a water body based on a deep confidence network. Background technique [0002] As aquaculture enters the stage of intensive cultivation, water quality management has become the most concerned issue of the aquaculture industry. The factors that affect water quality include biological, physical and chemical factors, such as water temperature, pH, water quality, ammonia nitrogen, and dissolved oxygen, among which water quality ammonia nitrogen is one of the main key indicators for detecting water quality. [0003] The ammonia nitrogen in the water body is an important part of the single cycle, which is biologically harmful to aquaculture objects. When it enters the aquatic organisms, it will cause breathing difficulties, decreased resistance, and non-eating, which will affect aquatic products. In severe cases, the q...

AI Technical Summary

Problems solved by technology

BP neural network can solve nonlinear and complex problems, but its training speed is too slow and it is too sensitive to external noise, resulting in poor stability and low accuracy of prediction results

Images