Electrical engineering

Physics

Heuristics

Sociology

Energy (signal processing)

Smart meter

Consumption (sociology)

Mathematics

Social science

Wireless

Biology

Automatic meter reading

Operating system

Statistics

Astronomy

Engineering

Heuristic

Metre

Paleontology

Artificial intelligence

Power (physics)

Electricity

Energy consumption

Context (archaeology)

Computer science

Quantum mechanics

Electricity meter

Telecommunications

Wireless Networks

Energy consumption is dependent on temperature, humidity, occupancy, occupant type, building area etc. All these factors collectively define the context of an energy meter. Once the context is known, the meters within the same context can be grouped and their behaviour can be analyzed together. This paper presents four heuristics, including one novel heuristic, to identify abnormal energy consumption. Using these heuristics, data collected from fifty smart meters deployed inside hostels of IIIT-Delhi was investigated for abnormal energy consumption detection. The anomalies and possible causes were discussed with IIIT-Delhi campus administrator. Energy consumption per occupant for one of the meters was found four times when compared to rest of the meters. The results demonstrated that the proposed heuristics successfully found abnormal energy consumption behaviour.

Detecting anomalous energy consumption using contextual analysis of smart meter data