A soft measurement approach of wastewater treatment process by lion swarm optimizer-based extreme learning machine

Some key variables in the wastewater treatment process, such as the biochemical oxygen demand (BOD5) and the chemical oxygen demand (COD), are difficult to perform timely and accurate measurement by hardware means. In view of this issue, we introduce a novel soft measurement approach. In the proposed approach, the sparse principal component analysis (SPCA) is used for dimensionality reduction of datasets, and construction of the soft measurement model is done by an improved extreme learning machine (ELM) algorithm, the lion swarm optimizer-based extreme learning machine (LSO-ELM). In the LSO-ELM, the connection weight matrix from the input layer to the hidden layer and the bias vector of the hidden layer are optimized by the LSO. The optimization of these two parameters has greatly improved the predictive ability of original ELM. We illustrate the LSO-ELM approach on Benchmark Simulation Model 1 (BSM1) and the results show that it has excellent performance on the prediction of BOD5 and COD. The proposed approach will greatly improve the quality of soft measurement in wastewater treatment process.

Automated summary

The proposed approach utilizes SPCA for dimensionality reduction and an improved ELM algorithm, LSO-ELM, for soft measurement model construction, showing excellent performance on BOD5 and COD prediction in wastewater treatment process.