Aquaculture dissolved oxygen prediction method and device

Abstract

The embodiment of the invention provides an aquaculture dissolved oxygen prediction method and device. The method includes the steps that water quality parameters and environment meteorological parameters of an aquaculture pond are acquired to serve as to-be-analyzed data, wherein the water quality parameters comprise the dissolved oxygen content; a principal component analysis method is adopted for carrying out dimension reduction and analysis on the to-be-analyzed data, principal components influencing the water quality dissolved oxygen change are acquired, and key influence factor data of the water quality dissolved oxygen is determined according to the principal components; the key influence factor data is input in a trained LSTM network model, and predicted data of the dissolved oxygen content of the aquaculture pond is obtained according to the output result of the LSTM network model. The problems that a traditional prediction model is low in prediction precision, poor in stability and the like are solved, the effect of efficiently predicting the water quality dissolved oxygen content is achieved, and a decision basis can be provided for regulation and control management of the aquaculture water quality dissolved oxygen.

Description

technical field [0001] Embodiments of the present invention relate to the fields of data mining and machine learning, and in particular to a method and device for predicting dissolved oxygen in aquaculture. Background technique [0002] With the rapid development of the aquaculture industry, it is of great significance to strengthen the research on key technologies of aquaculture water quality prediction for reducing the risk of aquaculture, preventing disease outbreaks and optimizing aquaculture management. Dissolved oxygen content is an important indicator reflecting the growth status of aquatic products and water quality. It is easily affected by various factors such as temperature, wind speed, wind direction, rainfall, aquatic organism metabolism and human activities, and has nonlinearity, large time lag and instability. Features. At present, the regulation of dissolved oxygen in aquaculture mainly relies on manual experience and weather to judge whether to increase oxy...

AI Technical Summary

Problems solved by technology

However, the single intelligent model has disadvantages such as easy to fall into local extremum, slow training speed, poor stability, and low prediction accuracy. The construction process of the combined model is complicated and artificially dependent, which is not conducive to popularization and use in practice

Images