GOMES, R. D.; http://lattes.cnpq.br/0944963449027456; GOMES, Ruan Delgado.
Resumo:
The optimization of energy usage, by decreasing costs and reducing environmental impact,
has been cause of great concern in various sectors of society. In this context, intelligent and
low-cost industrial automation systems for motor monitoring can be a very useful tool, allowing
reducing the use of electricity in the industrial sector. Traditionally, energy monitoring
and fault detection in industrial systems are performed off-line or through wired networks.
This latter approach has a high cost and limited flexibility, mainly due to the need of cables,
which difficult the process of installing and maintaining the network. An alternative to construction
of flexible and low-cost industrial monitoring systems is the use of Wireless Sensor
Networks (WSN). However, the application of these networks in an industrial environment
presents a number of challenges. Transmission errors and the variable link capacity are one
of the most critical factors in the application of WSN in an industrial environment. This work
proposes the development of a WSN for real-time monitoring of motors in industrial environment.
The network consists of intelligent nodes that perform local processing to estimate
the parameters (torque and efficiency), and transmit these values to a central monitoring unit.
Experimental studies were performed to observe the relationship between the WSN performance
and spectral occupancy in the operating environment, also observing the impact of
external interference sources. It was also developed a mathematical model to evaluate the
WSN performance in several scenarios. Through these theoretical and experimental studies,
it was demonstrated that employing intelligent nodes brings great benefits for this type of
application, reducing the amount of data transmitted over the network and allowing monitoring
even in high interference scenarios. In addition to that, our work provides.