Water body water quality prediction method, system and device

Abstract

The invention provides a water body water quality prediction method, system and device. The method comprises the following steps that: independently carrying out missing data processing and normalization processing on multiple groups of water quality parameters of a target water body and a corresponding meteorological parameter to obtain a standard meteorological parameter and a standard water quality parameter; and on the basis of the standard meteorological parameter and the standard water quality parameter, utilizing a trained autoencoder and a long short term memory network fusion prediction model to obtain the prediction water quality index of the target water body, wherein the long short term memory network fusion prediction model is obtained in a way that the standard meteorologicalparameter and the corersponding standard water quality parameter of the target water body are utilized in advance to train the constructed autoencoder and long short term memory network fusion prediction model in sequence. By use of the method, a mutual function relationship among water body multiple parameters can be effectively mined so as to effectively predict the water body water quality anddisclose the key rule of a water quality factor change.

Description

technical field [0001] The present invention relates to the technical field of information processing, and more specifically, to a water quality prediction method, system and device. Background technique [0002] With the growing scale of aquaculture, intensive aquaculture is gradually becoming the main aquaculture production method. The increase of breeding density, unscientific use of medicines and bait will damage the water quality, lead to frequent breeding diseases, and affect the economic benefits of aquaculture. On the other hand, the purification ability of the aquaculture water itself is poor, and the nitrogen, phosphate, and organic residues accumulated in the aquaculture process for a long time cannot be exchanged and diffused with the outside world, which easily causes pollutant deposition and eutrophication of the water body, polluting the ecological environment. [0003] Through the effective prediction of water quality, the state of aquaculture products can b...

AI Technical Summary

Problems solved by technology

However, the BP neural network optimized by the genetic algorithm is difficult to learn the interaction relationship between multiple parameters of the water body, and it is not enough to reveal the key laws of the change of water quality factors, especially the long-term dependence relationship between time data cannot be learned

Images