The objective of this project is to integrate some AI techniques (such as those based on neural networks, evolutionary computing and reinforcement learning), with advanced control algorithms (such as predictive control), to develop a control architecture capable of integrating easily in industrial control equipment that is commonly used in process control.
The design, development and validation of this new intelligent predictive control architecture, with the capacity to learn and generate robust models and capable of real-time control of industrial processes with complex dynamics, could be implemented in industrial control equipment used in the process sector chemical, mechanical or energetic, among others.

Industrial applicability

In fact, among the specific technological objectives of the project is the validation of the control algorithms developed in a fuel cell control system, an aluminum rolling process, a bridge crane control system or a control process of ph.

Equipment to validate results

The CPOH Group of the Institute of Automatic Control and Industrial Computing has a test bench for fuel cells in its laboratory, the result of previous projects, which will be used as a final testing platform for the developed architecture, not in simulation, but with an application finalist and realistic control of energy systems based on hydrogen cells.

With the aim of validating the algorithms, industrial control equipment with great computing capacity will also be used, which the CPOH has provided by various firms to the UPV laboratory.