•  I
• a · A  I
• Accesibility  I
• Map  I
• Search  I
• Staff directory

• ::  Iniciar sessió :: 

The researchers from the Universitat Politècnica de València (UPV) Miguel Enrique Iglesias Martínez, José Guerra Carmenate, José Alfonso Antonino Daviu, Larisa Dunai, José Alberto Conejero and Pedro Fernández de Córdoba have developed an innovative method that can detect faults in electric motors in industrial machines with great reliability and at very early stages using artificial intelligence. It can also be applied to electric vehicles. Their work, which was carried out in collaboration with Carlos A. Platero, from the Technical University of Madrid (UPM), has been awarded the prize for the best paper published in the field of electrical machines in 2023 in the IEEE Transactions on Industry Applications journal.

The method developed by the UPV and UPM team focuses on synchronous motors, which are essential in many industrial applications, especially those that demand very high power. These motors can reach several tons in weight and are often critical machines in the environments where they operate; their unexpected failures can cause millions of dollars in losses for companies due to unplanned stoppages of the production process or high repair costs.

‘The work proposes a new diagnostic method based on the analysis of dispersion flow signals, both in transient and steady state, using advanced signal processing techniques and artificial intelligence,’ explains José Alfonso Antonino, co-author of the study and professor at the Research Institute for Electrical Technology - UPV.

Easy to use, non-invasive and highly reliable

'One of the main contributions of this research is the use of scattering flow as a diagnostic tool, a magnitude that has attracted great interest among researchers and manufacturers due to its important advantages, such as its ease of use and non-invasiveness. In addition, the proposed method allows automatic fault diagnosis, i.e. without needing expert user intervention, which is ideal for implementation in stand-alone diagnostic systems,' adds Antonino.

The work applies a combination of modern signal analysis techniques based on the bispectrum and advanced image processing methods, 'which makes possible an automatic and highly reliable fault diagnosis', says Miguel Iglesias Martínez, a postdoctoral researcher in the Margarita Salas programme.

On the other hand, as José Alberto Conejero, professor at the Research Institute for Pure and Applied Mathematics - UPV, points out, "This research has had a markedly multidisciplinary character and combines the application of knowledge from different areas, such as electrical machines, signal processing and artificial intelligence".

Also for electric vehicles

Furthermore, according to the UPV team, the research has great potential ‘as it can be extrapolated to other areas in which electric motors play a fundamental role, such as marine propulsion, electric vehicles, electric planes or even robotics’, emphasises Pedro Fernández de Córdoba, also a professor at the Research Institute of Pure and Applied Mathematics.

The award will be presented during the annual meeting of the IEEE Industry Applications Society's Electrical Machinery Committee (EMC), within the framework of the IEEE Energy Conversion Congress & Expo (ECCE 2024), which will take place in Phoenix, USA, from 20 to 24 October.

This meeting brings together leading world experts in the field of Electrical Machines.