Author(s): K. A. Nguyen; H. Zhang; R. A. Stewart
Linked Author(s):
Keywords: Prototype selection; Dynamic Time Warping algorithm; Water end-use; Flow-trace disaggregation
Abstract: The disaggregation of domestic water consumption flow-trace data into end use event categories still remains a complex challenge to be resolved in the field of urban water management. Domestic end use studies currently utilise software and analyst experience to disaggregate flow data into end use events (e.g. faucet, dishwasher, toilet, etc.), which often requires in excess of two hours per home to disaggregate two weeks of flow data. An existing available database of end use events for over 200 households located in South-east Queensland (SEQ), Australia was utilised for the purpose of this study, which ultimately aims to automate the trace analysis process. One of the first research issues to be addressed in this research was to develop a prototype that encapsulates the wide variation in pattern characteristics for each end use event, in order to reduce unnecessary computation and memory resource required to analyse the entire end use database. To achieve this aim, the study employed the Dynamic Time Warping algorithm. The outcome of this practice is a series of prototypes representing the predominant domestic water end use events in a household (e.g. shower, clothes washer, etc.). Future work will employ the prototypes in an artificial network model to automate the end use disaggregation process. Moreover, validation of its accuracy will be examined through predicting end uses for a new sample of smart metered households.
Year: 2011