Multi-model Modeling for Activated Sludge Process Based on Clustering Analysis under Benchmark

For wastewater treatment processes, a single model suffers from heavy burden calculation and bad accuracy. A modeling method based on auto-regressive exogenous (ARX) multi-model using improved supervised k-means clustering algorithm is proposed. The method introduced the cluster center initialization idea of CCIA algorithm into classical k-means clustering algorithm applied to group the data into clusters or second clustering by judging a preset threshold value. It will improve the clustering results to make better services for the subsequent modeling work. And the least squares method is used to construct ARX sub-models. The system model is constructed by weighting all ARX sub-models. The proposed method is used to identify the ammonia concentration model for wastewater treatment system Benchmark. Simulation results show that, the proposed method can be used to fit nonlinear characteristics of the system with high precision.