Robust Adaptive Predictive Fault-Tolerant Control Integrated to a Fault-Detection System Applied to a Nonlinear Chemical Process

Most of the control schemes for chemical plants are developed under the assumption that the sensors and the actuators are free from faults. However, the occurrence of faults will cause degradation in the closed-loop performance, having an impact on safety, productivity, and plant economy. In this work, the main novelty is given by the enhancement produced through the integration of the fault detection and identification (FDI) system over a robust adaptive predictive control (RAPC) strategy specially thought to turn it as a fault- tolerant control (FTC) scheme. Additionally, the FDI itself is original because of the sensor and actuator faults treatment. The biases in sensors are detected and quantified by using wavelet decomposition and the extra delays in actuators by applying online identification techniques to appropriately modify the controller action. It is important to remark that the extra time delay, detected particularly at the actuators, is a problem that occurs frequently in practice; however, the academic community has mostly omitted it up to now. This methodology can improve the overall performance for nonlinear stable plants because the FDI is specifically designed as a complement of those aspects that RAPC cannot handle at all. The control technique involves a commutation of a linear time-varying robust filter in the feedback path of the control loop in synchronization with an adaptive predictive controller. Through simulation studies of a continuous stirred tank reactor (CSTR) with jacket, where the integration between FDI and FTC has been implemented, it can be shown that the proposed methodology leads to significant improvement in comparison with the same control scheme without FDI, particularly when the fault magnitude increases. Fil: Zumoffen, David Alejandro Ramon. Universidad Tecnológica Nacional; Argentina Fil: Basualdo, Marta Susana. Universidad Tecnológica Nacional; Argentina