Abstract
In an industrial environment, some phenomena such as wind, humidity, pressure or temperature are required to be controlled in a distributed and low invasive manner. This paper describes the development of a wireless sensor network that can easily be spread out and connected to a wide range of sensors. This versatility is reinforced by the processing and transmission capacities of each component of the network. An application to temperature mapping is provided to illustrate the functionality of the network. In particular, the previous hardware has been developed in order to perform temperature control while managing several airflow units. The learning based control proposal presented herein involves first, a neuro-fuzzy system in charge of generating a reliable model of the refrigeration process and then, an ant colony metaheuristic to find an optimum airflow solution that fits the required temperature distribution.
Keywords: Ant colony optimization, learning based control, neuro-fuzzy system, wireless sensor network.