Monitoring of energy consumption using Coded Power Radio

Reducing CO2 emission (enforced by legislation in some countries) and lowering the electricity bill are two valid reasons for consumers and businesses for wanting to reduce their electric energy consumption. The first step in this process is finding out how much energy is actually consumed by the various devices and appliances. A partial solution is the use of power meters that are inserted between the appliance and the socket. These allow for monitoring the energy consumption of individual devices. However, many power meters are needed, and an important part of the energy consumption (lighting, HVAC) cannot be monitored in this way.

An alternative approach, which has been successfully explored in the Coded Power project at Philips Research, is to measure the total electric current and voltage at a central location in the network, and to disaggregate the current into the contributions of individual appliances or groups of appliances. This appears to be feasible (at least to some extent), because each appliance, or class of appliances, has a characteristic current and/or power profile. A good disaggregation algorithm makes use of techniques from

• Signal processing
• Multi-user detection
• Event detection
• Pattern recognition
• Filtering
• Database management
• ...

Assignment

Philips Research is looking for an MSc student to help them in setting up a simulation environment for testing and combining various algorithms.

Requirements

The candidate is a Master's student in Electrical Engineering, with a solid signal processing and/or telecommunication background. Knowledge of Matlab is required. Good spoken and written English is a must. Expected project duration is 6-9 months.

The project is carried out at Philips Research, High Tech Campus 34, Eindhoven (ir. Ying Wang)

Contact

prof.dr.ir. Geert Leus

Signal Processing Systems Group

Department of Microelectronics

Last modified: 2017-10-20