EWMA based approaches for automated building energy analysis

With the introduction of automatic meter reading (AMR) technologies, high frequency data from building electricity, gas and water consumption is increasingly common. Large volumes of data are being generated and there is a growing demand for automatic analysis tools. This paper introduces two automatic analysis methods, based on exponential weighted moving average (EWMA) control charts. These two methods can identify changes in the trend of consumption during specific periods of the day, and changes in consumption relative to a mathematical model. The abnormal consumption events identified by these methods are in accordance with the events identified by expert energy analysts.