• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.5
• /
• pp.638-644
• /
• 2010

This paper presents ideas of service scenarios for home residents using electric power monitoring system per appliance, the implementation of the monitoring system, and analysis of acquired electric power usage pattern. By acquiring and analyzing electric power usage pattern, home residents can get information of power usage pattern of every legacy (non-Demand Response-ready) appliance. Further they can get pieces of recommendation how to reduce energy consumption, intelligent standby power blocking service, and alarming service to abnormality of appliances. In order to check the feasibility of the ideas, a system that can acquire electric power pattern per appliance is implemented, and electric power pattern of some appliances are stored to a database and it was analyzed to show if auto-identification of a type of a device is possible, which is a basic required function for the scenarios presented.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.4
• /
• pp.407-413
• /
• 2011

This paper presents an electrical power monitoring system for home energy management and an automatic appliance-identification algorithm based on the electricity-usage patterns collected during the monitoring tests. This paper also discusses the results of the field tests of which the proposed system was voluntarily deployed at 13 homes. The proposed monitoring system periodically measures the amount of power consumption of each appliance with a pre-specified time interval and effectively displays the essential information provided by the monitored data which is required users to know in order to save power consumption. Regarding the field tests of the monitoring system, the households responded that the system was useful in saving electricity and especially the electricity-usage patterns per appliances. They also considered that the predicted amount of the monthly power consumption was effective. The proposed appliance-identification algorithm uses 4 patterns: Zero-Crossing Rate(ZC), Variation of On State(VO), Slope of On State(SO) and Duty Cycle(DC), which are applied over the 2 hour interval with 25% of it on state, and it yielded 82.1% of success rate in identifying 5 kinds of appliances: refrigerator, TV, electric rice-cooker, kimchi-refrigerator and washing machine.