Smarter and greener energy technologies are expected replace the legacy power network in near future. Variety of innovative technologies, system architectures and market models are being investigated in order to significantly reduce the carbon footprint of the energy sector. This project investigates the advantages of active participation of customers in energy market through a customer friendly plug & play system architecture. Where, aggregated Distributed Energy Resources (DER) like solar PV, Electric Vehicles and controllable loads in customer domain are controlled in an optimal fashion to minimize carbon emission while ensuring customer satisfaction. In this project, in addition to developing the required optimization algorithm, the study was extended to model the effects of random communication delays on the aggregated output of DERs using applied probability theory.
In the proposed research, a system that can determine the dynamic line rating is developed using low cost solutions. GPRS communication is used to transmit data from sensors located on the power lines and these data and the parameters of the line will be used to determine the dynamic line rating. The sensors and communication unit will be powered by the electricity harnessed from the power line itself. This dynamic line rating system will be first implemented and tested in the laboratory. Then it is enclosed in a suitable enclosure and suitable field test are carried out on distribution overhead lines.
LKR. 2 million
2014 January – to date
Ceylon Electricity Board, Lanka Electricity Company.
National Science Foundation
This research is aiming to investigate how dc microgrid could help to reduce the losses in power distribution and be used to integrate PV into power system and how PV could enhance the quality, reliability and security of the power supply. Novel converter topologies related to dc networks will be investigated.
LKR. 2.2 million
02/06/2014 – 30/06/2017
University of Moratuwa
National Research Council, Sri Lanka
Load Monitoring techniques determine the appliances that are turned-on within a given period of time in a household. They play a critical role in a variety of smart grid applications such as supply and demand side power control, smart billing and intelligent appliance monitoring and control. Load monitoring can be performed both intrusively as well as non-intrusively.
Intrusive load monitoring estimates the turned-on appliances by attaching individual sensors to each appliance to be monitored. Non-Intrusive Load Monitoring (NILM) attempts to identify the turned-on appliances from the power supply entry point to the household or workplace. The necessity for effective and efficient NILM methods for residential appliance identification has recently escalated due to its application potential for smart grids. Most of the existing NILM methods require very high sampling rates to capture the unique features from the measurement signals. Further, some of these NILM methods require more than one electrical measurement (such as voltage, current, active power, reactive power etc.) for the appliances identification. Further, such methods need multi functional smart meters that are also costly. Considering the above drawbacks, a novel NILM method was proposed based on uncorrelated spectral information of a low frequency (less than 1 Hz) active power consumption signal. Real household active power consumption data from two public databases, i.e. tracebase and REDD, was used to demostrate the robustness of the proposed NILM method under several practical scenarios.
2014 May – to date
A key aspect of electricity supply quality in a power system is to supply voltages within its limits. In this research an electronically controlled Volt-Var Control (VVC) scheme based on a three winding transformer is investigated. Two configurations based on series and parallel compensation are studied for their relevant merits.
LKR. 2.25 million
2014 January – to date
Lanka Transformers Ltd.
InRC of University of Peradeniya, Lanka transformers Ltd.
Moisture estimation of transformer pressboard by Microstrip ring resonator at GHz frequencies
from October 2014
Collaboration between high voltage and microwave research groups.
|Classification Method||Results: Accuracy|
|Training Data||Test data|
|1. Simple linear boundary classification C4_RMS=C3_RMS||80%||80%|
|2. Voting using RMS||85%||85%|
|3. OPtimized linear boundary classification C4_RMS=m_opt*C3_RMS + C0_opt||82.9%||75%|
January 2014 – Aug 2015