Monitoring of wastewater treatment process based on multi-stage variational autoencoder

作者：

Highlights：

• Utilize the AP algorithm to divide the data of WWTP into several stages.

• The method is able to deal with the multi-stage problem of the WWTP.

• Establish a variational autoencoder process monitoring model in each sub-stage of the WWTP.

• The novel method can simultaneously process the non-linearity and non-Gaussian of the data.

摘要

•Utilize the AP algorithm to divide the data of WWTP into several stages.•The method is able to deal with the multi-stage problem of the WWTP.•Establish a variational autoencoder process monitoring model in each sub-stage of the WWTP.•The novel method can simultaneously process the non-linearity and non-Gaussian of the data.

论文关键词：Monitoring of wastewater treatment process,Multi-stage process,Non-linear process,Non-Gaussian process,Variational autoencoder

论文评审过程：Received 30 October 2020, Revised 20 May 2022, Accepted 17 June 2022, Available online 5 July 2022, Version of Record 11 July 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117919