کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
4962780 | 1446738 | 2017 | 9 صفحه PDF | دانلود رایگان |
- Result shows feasibility to use low cost off-the-shelf component to build an energy monitor.
- Satisfactory classification accuracies achieved by using few training examples with Naive Bayes classifier and one feature (average power).
- A data set developed consisting of 4 types of appliances from the application of embedded networking device.
- Two classification accuracies metrics obtained from the developed Naive Bayes classifier.
This paper reports the use of low cost off-the-shelf components to develop a wireless energy monitoring system for the purpose of non-intrusive appliance load monitoring. The system comprises hardware design where a standard open source WiFi-connected embedded system board is used as the core component, an analog front end for interfacing between the voltage and current sensors and the embedded board and also a server to store data for further processing. The software side consists of algorithms for performing data partitioning and Naive Bayes classifier. Non-intrusive appliance load monitoring (NIALM) is a methodology used to disaggregate total power consumption into individual electrical appliance power usage. In this paper, the average power consumptions of electrical appliances are obtained from the system setup through the identification of switching ON or OFF events of appliances. The Naive Bayes classifier is deployed for appliances' classification. This work demonstrates the feasibility of utilizing low cost and standard components to develop a wireless energy monitoring system. It is also observed that the developed classifier algorithm is able to classify individual electrical appliances with satisfactory accuracy level using few training examples. The novelty of this paper is that it developed a data set consisting of 4 types of appliances from the application of embedded networking device. This paper also shows that satisfactory classification accuracies can be achieved using few training examples with generic classifiers (Naive Bayes) and one feature (the average power). It provides an example of how NIALM can offer feedback to appliances' energy consumption and therefore promote energy efficiency.
Journal: Sustainable Computing: Informatics and Systems - Volume 14, June 2017, Pages 34-42