Neural networks for process parameter identification and assisted controller tuning for control loops

The paper introduces a method for identification and assisted controller tuning for industrial process control loops suitable for an 'in-chip' solution. The method is based on utilization of neural network technology. It employs pattern recognition and exploits the nonlinear function approximation feature of neural networks. The method has the advantage of simplicity and sufficient accuracy. It is developed for on-line usage in closed-loop operations