A hybrid model of empirical wavelet transform and extreme learning machine for dissolved oxygen forecasting
by Juan Huan; Weijian Cao; Yuwan Gu; Yilin Qin
International Journal of Embedded Systems (IJES), Vol. 13, No. 1, 2020

Abstract: The accurate predicting trend of dissolved oxygen (DO) can reduce the risks to aquaculture, so a combined nonlinear prediction model based on empirical wavelet transform (EWT) and extreme learning machine (ELM) optimised by adaptive disturbance particle swarm optimisation (ADPSO) is proposed. First of all, DO series are decomposed into a term of relatively subsequence by EWT, secondly, the decomposed components are reconstructed using the C-C method, and thirdly an ELM prediction model of every component is established. At last, the predicted values of DO datasets are calculated by using RBF to reconstruct the forecasting values of all components. This model is tested in the special aquaculture farm in Liyang City, Jiangsu Province. Results indicate that the proposed prediction model of EWT-ELM has better performance than WD-ELM, EMD-ELM, ELM and EWT-BP. The research shows that the combined forecasting model can effectively extract the sequence characteristics, and can provide a basis for decision-making management of water quality, which has certain application value.

Online publication date: Wed, 08-Jul-2020

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Embedded Systems (IJES):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?


Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com