DeepEnergy: Prediction of Appliances Energy with Long-Short Term Memory Recurrent Neural Network

Abstract

Our world is becoming more interconnected and intelligent, huge amount of data has been generated newly. Home appliances’ energy usage is the basis of home energy management and highly depends on weather condition and environment. Using weather in context, it is theorized that usage of home energy would be higher in cold days. Time series and contextual data collected from sensors can be monitored and controlled in home appliances network. The aim of this work is to propose a deep neural network architecture and apply it to a contextual and multivariate time series data. Long short-term memory (LSTM) models are powerful neural networks based on past behaviours in long sequences. LSTM networks have been demonstrated to be particularly useful for learning sequences containing longer-term patterns of unknown length, due to their ability to maintain long-term memory. In this work, we incorporate contextual features into the LSTM model because of ability of keeping context of data for a long-time, and for analysing it we integrated two different datasets; the first dataset contains measurements about house temperature and humidity measured over a period of 4.5 months by a 10 min intervals using a ZigBee wireless sensor network. The second dataset contains measurements about individual household electric power consumptions gathered over a period of 47 months. From the wireless network, the data from the kitchen, laundry and living room were ranked the highest in importance for the energy prediction.