Fault detection for wastewater treatment bioprocesses based on ellipsoid bundles

This paper proposes a fault detection approach for the basic process of microbial growth in wastewater treatment, based on interval estimation technology. First, a nonlinear microbial growth model is converted into a Takagi–Sugeno (T–S) fuzzy system. Then, utilizing a T–S fuzzy structure, an observer is constructed based on an $L_{\infty }$ performance index. Furthermore, an ellipsoid bundle based set membership estimation algorithm is analyzed to synchronize interval estimation operations. In addition, interval residuals are computed from the generated interval state estimation and fault detection is performed for a wastewater treatment bioprocess with actuator faults or sensor faults. Finally, the feasibility of the proposed fault detection strategy for the wastewater treatment bioprocess is illustrated through numerical simulations.