A comparative study on a novel model‐based PID tuning and control mechanism for nonlinear systems

Abstract

SUMMARYThis work presents a novel predictive model-based proportional integral derivative (PID) tuning and control approach for unknown nonlinear systems. For this purpose, an NARX model of the plant to be controlled is obtained and then it used for both PID tuning and correction of the control action. In this study, for comparison, neural networks (NNs) and support vector machines (SVMs) have been used for modeling. The proposed structure has been tested on two highly nonlinear systems via simulations by comparing control and convergence performances of SVM-and NN-Based PID controllers. The simulation results have shown that when used in the proposed scheme, both NN and SVM approaches provide rapid parameter convergence and considerably high control performance by yielding very small transient-and steady-state tracking errors. Moreover, they can maintain their control performances under noisy conditions, while convergence properties are deteriorated to some extent due to the measurement noises.

B-Spline Neural Networks Based PID Controller for Hammerstein Systems

“…Note also that the calculation of multistep ahead predictions and the Jacobian J m are model specific. In [11], this controller scheme has been applied to unknown nonlinear systems that are modelled using neural networks and support vector machines (SVMs) respectively. In the following, the calculation of multistep ahead predictions and Jacobian J m for the B-spline neural network based Hammerstein model are introduced.…”

“…Neural networks have been widely applied to model unknown dynamical processes and then used for PID parameter tuning [13,6,16]. Recently a novel predictive model-based PID tuning and control approach has been proposed for unknown nonlinear systems that are modelled using neural networks and support vector machines (SVMs) [11]. The work introduces a useful technique both for PID parameter tuning and for the correction of the PID output during control, which yields superior tracking and parameter convergence performance.…”

“…For system identification we used the Gauss-Newton algorithm subject to constraints as proposed in [9]. The predictive model-based PID tuning and controller approach in [11] was combined with the B-spline neural network based Hammerstein model. For this purpose, multistep ahead predictions of the B-spline neural networks based Hammerstein model are generated as well as the essential Jacobian matrix for updating the control signal, based on the De Boor recursion including both the functional and derivative recursions.…”

A model‐based PID controller for Hammerstein systems using B‐spline neural networks

“…The motivation of the proposed methods is twofold. First, this extends the model-based PID controller [12] to accommodate the Hammerstein systems. Second, the proposed model based on the B-spline neural networks has a significant advantage over many other modeling paradigms in that this enables stable and efficient evaluations of functional and derivative values on the basis of the de Boor recursion, which is used in updating the PID control signals.…”

“…In this paper, we used the Gauss-Newton algorithm subject to constraints, as proposed in [31]. The predictive model-based PID tuning and controller approach in [12] was combined with the B-spline neural network-based Hammerstein model. For this purpose, multistep ahead predictions of the B-spline neural network-based Hammerstein model are generated as well as the essential Jacobian matrix for updating the control signal, on the basis of the de Boor recursion including both the functional and derivative recursions.…”