Monday, June 24, 2013

Tales from PowerTech 2013 in Grenoble

The PowerTech Conference is the main conference of the IEEE Power Engineering Society in Europe. At least this is what we were told in the opening speech on Monday the 17th of June in the World Trade Center of Grenoble. And honestly they really made a lot of effort to leverage the conference impact. Even though the French Minister for research and higher education did not made it personally to the sight still her video message was part of the conference opening session. Executive officers from local transmission and distribution grid operators and electric industries had a slot in that session as well. You might have guessed, by help of the picture, that one of them was from Schneider Electrics which is a quite important player in that region. But there is even more famous things coming from Grenoble: It was the town where Fourier was working and teaching. His house is accessible for public as a museum. Unfortunately I could not find time to visit it as many interesting talks waited to be listened to. 

The representers of industries gave an idea about how their companies deal with the topic Smart Grid. The speakers associated with political institutions gave their perspective about the problems to be solved. It was mentioned several times in this session that there is a big leak of business models for the smart grid. This is associated with missing motivation to invest in technology and its general roll out. Further I was quite surprised to see very familiar content in the slides of the invited speaker M. Masera from the JRC-European Commission. He presented slides related to Ettore Bombard's presentation from last year's Lakeside Labs Research Days about modeling a social layer for smart grid, including user behavior. My own presentation titled “Simulating the Smart Grid” was scheduled to the second day right after the invited speakers in the session about dynamic modeling:

  • M. Poechacker, A. Sobe, and W. Elmenreich. Simulating the smart grid. In Proceedings of IEEE PowerTech, Grenoble, France, 2013.

About seven to eight different tracks were running simultaneously over the 11 sessions, additionally two poster sessions, several invited speakers and special sessions held by companies. Around 500 attendees from all over the world, mainly Europe, have been there. 
The main conference program was enriched by technical visits to power system related facilities and the touristic visits in the region. I participated at a tour to the INES (Institute Nacional de Energia Solar), an impressive research center with around 400 employees. The visit has good chance to become content of a further post in this blog. I would say the organizers did a good job. A resume that is strong influenced by the daily served lunch (including the obligatory French chees and wine) and the nice Gala dinner (with life music and artistic presentations) in the ice stadium of Grenoble. 

If I caught your interest in this conference now, the next PowerTech conference will be 2015 in Eindhoven, Netherlands.

Monday, June 17, 2013

Evolutionary Appliance Detection for Miscellaneous Household Appliances

The paper "EvoNILM - Evolutionary Appliance Detection for Miscellaneous Household Appliances" was accepted to the Workshop "Green and Efficient Energy Applications of Genetic and Evolutionary Computation" at Gecco 2013.

To improve the energy awareness of consumers, it is necessary to provide them with information about their energy demand, not just on the household level. Non-intrusive load monitoring (NILM) gives the consumer the opportunity to disaggregate their consumed power on the appliance level. The consumer is provided with information about the energy demand of each individual appliances. In this paper we present an evolutionary optimization algorithm, applicable to NILM purposes. It can be used to detect appliances with a probabilistic power demand model. We show that the detection performance of the evolutionary algorithm can be improved if the single population approach of the evolutionary algorithm is replaced by a parallel population approach with individual exchange and by the introduction of application-oriented pre-processing and mutation methods. The proposed algorithm is tested with Matlab simulations and is evaluated according to the fitness reached and detection probability of the algorithm.

This paper is an improvement and follow up paper of the previous work "Evolving Non-Intrusive Load Monitoring".