Algorithms help manage power loads
Two teams from the Lucerne University of Applied Sciences have worked with companies and an electricity company to develop a system that helps electricity suppliers postpone consumption peaks. The prototype can save you and your customers money.
The simultaneous activation of many electrical devices creates enormous network loads. Since up to 60 percent of the grid usage costs for the energy supply companies are calculated from these load peaks, they have a vital interest in avoiding or at least smoothing them out as far as possible. The Lucerne University of Applied Sciences and Arts ( HSLU ) has now found a solution together with partners, according to a press release.
What is needed is “an intelligent and efficient energy and load management, with the help of which energy can be shifted”, Andrew Paice is quoted in this communication . He is the head of the HSLU Building Intelligence Research Center called iHomeLab . His team developed the prototype of such a system together with the HSLU Competence Center for Thermal Energy Storage, the Vilters-Wangs electricity company and business partners Asgal Informatik and Semax . The Swiss Agency for Innovation Promotion Innosuisse supports the project.
The artificial intelligence developed by the teams and its algorithms use the data from intelligent electricity reading devices, so-called smart meters. The researchers used it to calculate thermal models of buildings. The algorithms “identify power-consuming devices such as heat pumps, boilers or e-mobiles and power-producing devices such as photovoltaic systems,” explains Paice. Then maximum power consumption, switch-on and switch-off times and energy consumption per day are combined with temperature and weather forecasts. This enabled precise predictions to be made.
With these project results, the two companies involved in the project could now offer energy suppliers a service that would help them save costs – without the need for additional investments in their distribution networks.