Raghunath Reddy received his doctorate in Computer Science and Engineering (CSE). His research work was supervised by Prof. Vishal Garg. Here’s a summary of his research work on Robust load identification algorithms under varying voltages:
A rise in energy demand is a major cause of global warming and climate change. The U.S. Energy Information Administration (EIA) projects that the world energy consumption will grow by nearly 50%, and the energy consumed in the building sector will increase around 65% by 2050. Therefore, two crucial sustainability measures are to improve energy efficiency and reduce energy consumption. To assist in sustainability measures, Appliance Load Monitoring (ALM) is essential for effective load management because it determines energy consumption and operating states of individual appliances. Evidence shows that energy feedback information can enable consumers to reduce consumption between 5% and 15% and that appliance-specific consumption information is more useful than aggregate information. Most electrical appliances do not have the capabilities to monitor their individual energy consumption details. To estimate this energy consumption for appliances, the two most widely used load monitoring approaches are Intrusive Load Monitoring (ILM) and Non-Intrusive Load Monitoring (NILM). ILM makes use of smart power strips or plug sockets to monitor appliances. NILM or load disaggregation is an approach to estimate individual appliance energy consumption from aggregate load measurement obtained from a single measurement point. Features extracted from electrical signals are used for automatic appliance identification. Machine learning-based ILM and NILM techniques have been published by researchers in the past. However, these machine learning-based ILM/NILM techniques are sensitive to noise, and their performance degrades in the presence of variations in data due to noise, sensor drift, source voltage fluctuations, and so on. In developing countries, the voltage fluctuates from 210 to 245 volts. The ILM and NILM approaches are not tested in these varying operating conditions. There is a need for developing a low-cost, efficient, and accurate load monitoring solution. This thesis aims to investigate the robustness of supervised load identification techniques for improving ILM and NILM approaches. The first part of the thesis provides techniques to improve ILM approach especially to monitor plug loads. Plug loads are the appliances that are plugged in to AC sockets. Plug loads account for 20% to 30% of building energy consumption and there is an increasing trend in plug load consumption. The energy performance of buildings can be improved by effective plug load monitoring and control. We conducted a plug load monitoring field study in an academic institute to gain insights into plug load usage and energy consumption patterns. Large scale deployment of such load monitoring solutions is costly. Without load-sensing capabilities, it isn’t easy to track some of the plug loads that keep changing locations. Smart plug strips or smart sockets with plugged in load identification capability are useful to automatically monitor loads. Smart strips or sockets record electrical measurements using an energy metering sensor. The extracted features or load signatures are used to train machine learning models for load identification. We provide a comparative performance analysis of these load identification techniques. The results show that a simple algorithm like KNN on low-frequency data of two minutes provides excellent identification accuracy. However, the identification models had problems in the presence of voltage variations. A novel Regression-based Nearest Neighbour (RBNN) Classifier is developed for robust plug load identification under varying voltages. We use regression to recognize the behavior of plug loads based on their energy consumption signature and then use identified behavior for classification. We evaluated the proposed algorithm against the standard classification algorithm on a dataset of 70 plug loads operating in varying voltage conditions and experimental results show that the proposed algorithm performs better than standard classifiers in most cases. We have also designed a prototype smart power strip with the proposed load identification technique. The remaining part of this thesis describes ways to improve event-based NILM techniques. An eventbased NILM approach identifies ON or OFF events and extracts features to build appliance identification models. We study how an event-based NILM algorithm performs under the influence of varying voltages. We suggest feature extraction and feature selection techniques for reducing the impact of voltage variation. Appliance load identification methods in NILM generally use either steady-state or transient features for load identification. We hypothesize that these are complementary features, and so a hybrid combination of them will result in an improved appliance signature. A feature fusion-based NILM technique for appliance identification is developed. We design low dimensional hybrid features for appliance identification using Naive Bayes, K-Nearest Neighbor, Decision Trees, and Random Forest classifiers. The proposed NILM methodology is evaluated for robustness in changing environments. Experimental results show that our proposed feature fusion-based algorithms are more robust and outperform steadystate and transient feature-based algorithms by at least 9% and 15%, respectively.