Soft Sensor Modeling of Sewage Disposal Process Based on 　Multi-model Fuzzy Kernel Clustering Method

The combination of multi-models can improve the accuracy and robustness of model. Based on the above ideas, this paper proposes a multi-model fuzzy kernel clustering method. By using fuzzy kernel clustering, the input data are firstly clustered according to different work conditions. Then, the submodel is built for every subset by using least square support vector machine method. Finally, the system output is obtained through the sub -model switch strategy. The simulation on the platform of sewage disposal is made for the soft sensor model of biochemical oxygen demand (BOD) and some better results are obtained.