Artificial neural networks (ANNs) were built to predict coagulant (Model I) and alkalizer (Model II) dosages given raw and treated water parameters from a water clarifying process. Different ANN architectures were tested and optimal results were obtained with [

No abstract

This paper presents an application of a stabilizing model predictive control (IHMPC) strategy with the underlying guarantee of feasibility to an oil production well system with Electric Submersible Pump (ESP) installation. The proposed controller is compared with a conventional finite-horizon MPC which provides some unfeasible solutions in the presence of an unmeasured disturbance, due to the typical conflict among the ESP-lifted oil well system constraints. The results show IHMPC as a viable choice to improve ESP-lifted well production since it can incorporate the desired requirements and provides stabilizing control actions.

No abstract

No abstract