Web Processing Control using Backstepping and RBF Neural Networks

A sliding mode controller integrated moment of inertia observer of web handling systems

“…However, the PID controller lacks robustness and adapts to nonlinear models with uncertainty and changes in model parameters. For this problem, nonlinear control algorithms under the backstepping technique are a solution developed by many researchers and have had specific results [8,9]. In [10,11], the authors have designed a controller that is a perfect combination of the backstepping technique and a modified genetic algorithm (GA).…”

Model‐free hierarchical control with fractional‐order sliding surface for multisection web machines

“…The authors innovated a Backstepping-based Control structure incorporated with an RBF network for roll inertia change estimation. 45 An RBF Neural Network-based Backstepping Sliding Mode Control (RBFNN-BSMC) 46,47 in which changeable inertia of rolls is approximated for the adaptive ability of control system was addressed to reach tracking and flexibility goals. Also, an equivalent control architecture was suggested as in the study, 48 of which remarkable innovation is to estimate the derivative of virtual control signal such that the "explosion of terms" phenomenon is successfully eliminated.…”

“…[22][23][24][25][26] Apart from that, the use of fractional-order sliding surface plays a vital part in manually optimizing system responses, and hence system performance is improved compared to using traditional sliding surfaces as shown in the works. [27][28][29][30] (3) Instead of just approximating the moment of inertia in the study 45 or calculating uncertainties and disturbances by using fuzzy logic in the study, 30 an RBF neural network-based robust controllers (RBFNN-BRC), which is capable of adapting to uncertainties and disturbances, is implemented to estimate the complete model. Subsequently, the whole system is also easily proven stable by the Lyapunov stability theory.…”

“…Unfortunately, unpredictable changes in the moment of inertia of web winding systems make it difficult to specify an exact dynamic model in detail. The authors innovated a Backstepping‐based Control structure incorporated with an RBF network for roll inertia change estimation 45 . An RBF Neural Network‐based Backstepping Sliding Mode Control (RBFNN‐BSMC) 46,47 in which changeable inertia of rolls is approximated for the adaptive ability of control system was addressed to reach tracking and flexibility goals.…”

“…(3) Instead of just approximating the moment of inertia in the study 45 or calculating uncertainties and disturbances by using fuzzy logic in the study, 30 an RBF neural network‐based robust controllers (RBFNN‐BRC), which is capable of adapting to uncertainties and disturbances, is implemented to estimate the complete model. Subsequently, the whole system is also easily proven stable by the Lyapunov stability theory.…”