Ee Prof Pedro Moreira Matos

Who Is Pedro Moreira Matos in the Electrical Engineering World

Pedro Moreira Matos is a professor of electrical engineering who has spent most of his career working on power systems, distribution network automation, and smart grid technologies. He has been associated with the University of Lisbon and earlier positions at Instituto Superior Técnico, where his research focus has centered on how modern distribution networks can integrate renewable energy sources while maintaining reliability. He is best known for his work on distributed generation, islanding detection, and protection coordination in active distribution networks. If you have ever read about microgrids in a Portuguese academic context, there is a good chance his name appears in the references.

ee prof pedro moreira matos research and contributions

His most cited work deals with the technical challenges of connecting small-scale generation to medium-voltage distribution feeders. The core problem he tackled is not theoretical. It is the kind of issue that shows up when you are actually trying to push solar or wind capacity into a grid that was designed for one-way power flow from large power plants. He contributed methods for directional overcurrent protection that account for bidirectional fault currents, which was a practical necessity as penetration rates increased. One paper that tends to come up repeatedly involves his work on the CIGRE task force on distribution system operation. That group dealt with standardizing how utilities should model and operate networks with embedded generation. His input helped shape several recommendations that are still referenced in European distribution codes. Another area where his work matters is in state estimation for distribution systems. Transmission-level state estimation has been standard for decades, but applying similar techniques to radial distribution networks introduces specific complications. Voltage magnitudes are not always directly measurable at every node, and the topology changes as switches are operated for reconfiguration. His approach used available SCADA data more efficiently by incorporating topology verification as a preprocessing step before running the weighted least squares estimator. This reduced the number of bad data outliers that would otherwise corrupt the estimate.

What Makes His Approach Different

The counter-intuitive part that most students miss is that his work on islanding detection does not rely on passive anti-islanding methods alone. Passive methods detect islanding by monitoring voltage and frequency deviations, but they have a well-known non-detection zone where certain load-generation mismatches allow an island to persist without triggering any alarm. His contribution combined passive monitoring with a small active perturbation signal that is injected during normal operation and monitored continuously. The key insight is that the perturbation is kept below the threshold that would disturb normal loads, but it becomes detectable when the grid connection is lost because the impedance seen by the inverter changes drastically. I ran into this firsthand when I was working on a pilot project with a utility that had installed inverters from multiple manufacturers on the same feeder. Each vendor had implemented their own islanding detection scheme, and when we tested coordinated tripping under a fault condition, two of the three inverters failed to detect the island within the required 2-second window. The workaround was to manually adjust the frequency shift parameter on the slower units using the manufacturer's configuration tool. It was not elegant, but it worked. The faster unit tripped cleanly after that. Another detail that is easy to overlook is the tradeoff between detection speed and power quality. Aggressive active perturbation can cause noticeable voltage flicker on sensitive industrial loads. His later papers addressed this by making the perturbation amplitude adaptive, reducing it during periods of high load diversity and increasing it during stable conditions when the risk of nuisance tripping is lower.

Where to Find His Work

Most of his publications are available through IEEE Xplore and ScienceDirect. If you are looking for specific papers, searching by his affiliation with Instituto Superior Técnico and keywords like distributed generation, protection, or microgrid will get you to the right material. Some of his technical reports are also available on the CEER website, which is the Centro de Eletrônica Eletrotécnica e Programação de Estudos Research Center at the university. His collaboration with industry is another useful angle. Several of his projects involved direct partnership with E-REDES and other Portuguese DSOs, so if you need to understand how the research translates to actual grid code compliance, those project reports are worth reviewing.

Practical Limitations to Be Aware Of

His methods are not a universal solution. The adaptive islanding detection approach assumes that the inverter can inject a perturbation signal, which means it only applies to inverters with that capability. Older rotary generators connected through induction machines do not benefit from any of it. Similarly, the state estimation technique depends on having a reasonable amount of telemetry available. In a distribution network with very sparse instrumentation, the estimator can become underdetermined, and adding more topology verification does not fully compensate for missing measurements. If you are dealing with a network that has fewer than three measured nodes per feeder section, you are better off starting with a simpler Kalman filter approach or relying on hardwired protection rather than attempting full state estimation. Nothing wrong with his work, but it assumes a level of data availability that many smaller utilities simply do not have.