Sensor fault detection and isolation for chemical batch reactors

In this paper a scheme for detection and isolation of sensor faults in chemical batch reactors is proposed. The scheme is based on a bank of two observers for residual generation which guarantees sensor fault detection and isolation in presence of external disturbances and model uncertainties. In the observers a Hog approach is adopted for the design of the gains, while the unknown dynamics of the reactor (i.e., the heat released by the reaction) are estimated by an on-line interpolator based on a radial basis functions (RBF) neural network. Finally, the estimates provided by the observers and the sensor measures are processed by a decision making system (DMS) that provides information about the faulty sensor and an healthy measure. In order to test the effectiveness of the proposed approach, a simulation case study is developed