Walter Scott, Jr. College of Engineering

Graduate Exam Abstract

Sachin Soman
M.S. Final
May 22, 2014, 10am
Powerhouse Board Room 342
Analysis and Charecterization of Wireless Smart Power Meter
Abstract: Recent increases in the demand for and price of electricity has stimulated interest in monitoring energy usage and improving efficiency. This research work supports development of a low-cost wireless smart power meter capable of measuring VRMS, IRMS, real power, and reactive power. The proposed smart power meter features include matching by-device rate of consumption and usage patterns to assist users in monitoring the connected devices. The meter also includes condition monitoring to detect harmonics of interest in the connected circuits which can give vital clues about the defects in machines connected to the circuits. This research work focuses on estimating communicational and computational requirements of the smart power meter and optimization of the system based on the estimated communication and computational requirements.
The wireless communication capabilities investigated here are limited to existing wireless technologies in the environment where the power meters will be deployed. Field tests are performed to measure the performance of selected wireless standard in the deployment environment. The test results are used to understand the distance over which the smart power meters can communicate and where it is necessary to utilize repeaters or range extenders to reduce the data loss.
Computational requirements included analysis of smart meter front-end sampling of analog data from both current and voltage sensors. Digitized samples are stored in a buffer which is further processed by a microcontroller for all the desired results from the power meter. The various stages for processing the data require computational bandwidth and memory dependent on the size of the data stream and calculations involved in the particular stage. A Simulink-based system model of the power meter was developed to report a statistic of computational bandwidth demanded by each stage of data processing.
The developed smart meter works in an environment with other wireless devices which include Wi-Fi and Bluetooth. The data loss caused when the smart power meter transmits the data depends on the architecture of the wireless network and also pre-existing wireless technology working in the same environment and while operating in the same frequency band. The best approach in developing a wireless network should reduce the hardware cost of the network and to reduce the data loss in the wireless network. A wireless sensor network is simulated in OMNET++ platform to measure the performance of wireless standard used in smart power meters. Scenarios involving the number of routers in the network and varying throughput between devices are considered to measure the performance of wireless power meters.
Supplementary documents provided with the electronic version of this thesis contain program codes which were developed in Simulink and OMNET++.
Adviser: Dr. Peter Young
Co-Adviser: N/A
Non-ECE Member: Prof. Dan Zimmerle
Member 3: Dr. Sudeep Pasricha
Addional Members: N/A
Program of Study: