Fault-tolerant PID controllers design for unknown nonlinear systems based on support vector machine

PID controllers have been widely used in many industries. They can provide robust and reliable performance for most systems. However, it is well known that the automatic controlled systems are susceptible to faults and the faults can cause a malfunction of the loop. The main motivation in this paper is to present a design scheme of controllers using the least squares support vector machines (LS-SVM) which achieve to self-tune the parameters to compensate effects of the faults. This method used LS-SVM to learn system post-fault dynamics. The auto-tuning algorithm for PID controller is activated using the residual signal generated from the fault detection path. That is derived with the Lyapunov method and therefore, the model predicted tracking error is guaranteed to converge asymptotically. The strategy is clarified with the aid of a well-regarded benchmark three-tank system with various faults.