Nauman Zaffar

Support Vector Machine based fault detection & classification in smart grids

Smart Grids have recently attracted the attention of many profound research groups with their ability to create an automated and distributed energy level delivery. Computational Intelligence (CI) has been incorporated into various aspects of the smart grids, including fault detection and classification, which is a key issue in all the power systems. This paper presents two novel techniques for fault detection and classification in power Transmission Lines (TL). The proposed approaches are based on One-Class Quarter-Sphere Support Vector Machine (QSSVM). The first technique, Temporal-attribute QSSVM (TA-QSSVM), exploits the temporal and attribute correlations of the data measured in a TL for fault detection during the transient stage. The second technique is based on a novel One-Class SVM formulation, named as Attribute-QSSVM (A-QSSVM), that exploits attribute correlations only for automatic fault classification. The results indicate a detection and classification accuracy as high as 99%. Significant reduction (from O(n4) to O(n2)) in computational complexity is achieved as compared to the state-of-the-art techniques, which use Multi-Class SVM for fault classification. Moreover, unlike state-of-the-art techniques, both of these techniques are unsupervised and online and can be implemented on the existing monitoring infrastructure for online monitoring, fault detection and classification in power sytems.

Object segmentation using feature based conditional morphology

This paper presents a new technique to segment objects of interest from cluttered background with varying edge densities and illumination conditions from gray scale imagery. An optimal background model is generated and an index of disparity of the objects from this model is computed. This index estimates the disparity, both in terms of edge densities and edge orientation. We introduce feature based conditional morphology to process the representations that are most likely to belong to the object of interest and obtain a distilled edge map. These edges are linked using N/sup th/ order interpolation to get the final outline of the object. We compare our approach with 9 contemporary background subtraction algorithms (Toyama et al. (1999)). Our approach shows significant performance advantages and uses only the gray scale images, while the other approaches also need the color images for their algorithms. A comparison with the conventional morphological techniques is also made to highlight the advantages of our algorithms.

Assessment of harmonic pollution by LED lamps in power systems

Modern energy-efficient LED lamps draw non-sinusoidal currents from the grid. The harmonic distortion level varies with the quality of lamps non-linear driver circuit. This paper studies the problem of system-wide harmonic impact with consideration to both the variation in lighting load and difference in quality of LED lamps available in the market. An electrical classification model is proposed for LED lamps based on their harmonic spectrum and an analysis is carried out to quantify the distortion and losses possible in power system if such lamps are to be installed. Extending prior work in this area, we use advanced harmonic analysis techniques and conduct evaluation with voltage-dependent harmonic current model for LED driver. We built a harmonic analysis tool using Gauss-Seidel Method to carry out our evaluations on IEEE-30 test bus system. We report our results using IEEE standard 1459 which provides definitions of power quantities under non-sinusoidal conditions. Evaluations show that poor quality LED lamps can cause significant distortion in a power system and good quality LED lamps introduce significantly lower distortion while providing high economic and environmental benefits.

A Comparative Analysis of Electrical and Photo Characteristics of LED Lights

An increasing shift from conventional incandescent lamps to CFLs followed by LED lights has taken place over the last few years. LED lights promise 80% energy conservation over existing lighting infrastructure and to a whopping 95% energy savings compared to the incandescent light bulbs. This paper presents a study conducted by authors on energy efficient lighting solutions in Pakistan. The study includes measurements of power and photo characteristics of a variety of LED lights obtained from international and local vendors along with an analysis and comparison with IEC standards. The work established a test bed to measure power and photo characteristics of different type of lights for comparative analysis. The results from measurements reveal a need of proactive regulatory authority to put restrictions on power characteristics and quality of available local lights. The work also demonstrates photo characteristics difference in local and international LED lights.

Solar PV-Based Scalable DC Microgrid for Rural Electrification in Developing Regions

In this paper, we detail the design, analysis, and implementation of a highly distributed off-grid solar photovoltaic dc microgrid architecture suitable for rural electrification in developing countries. The proposed architecture is superior in comparison with existing architectures for rural electrification because of its 1) generation and storage scalability, 2) higher distribution efficiency (because of distributed generation and distributed storage for lower line losses), 3) ability to provide power for larger communal loads without the requirement for large, dedicated generation by extracting the benefit of usage diversity, and 4) localized control by using the hysteresis-based voltage droop method, thus eliminating the need for a central controller. The proposed microgrid architecture consists of several nanogrids capable of the self-sustained generation, storage, and bidirectional flow of power within the microgrid. Bidirectional power flow and distributed voltage droop control are implemented through the duty cycle control of a modified flyback converter. A detailed analysis in terms of power flow, loss, and system efficiency was conducted by using the Newton–Raphson method modified for dc power flow at varying distribution voltages, conductor sizes, and schemes of interconnection among the contributing nanogrids. A scaled-down version of the proposed architecture with various power sharing scenarios was also implemented on hardware, and yielded satisfactory results.

Networked control of single-phase induction motor air conditioners for constrained power systems

Split-Type Air Conditioners (ACs) are extensively used in households, commercial organizations and industrial setups. The duty-cycle based hysteretic control is inefficient and requires high power provisioning of full startup current for the compressor at each turn-on. This is especially demanding in constrained power systems with weak grids where people are dependent on backup generators and UPS for power needs during load-shedding. Some work has been done on optimizing energy consumption of Split-type ACs based on DC Compressors e.g. permanent magnet brushless DC motors (PMBLDCM) and three-phase induction machines but limited work has been done on operational optimization and energy efficiency of Split-type ACs based on single phase permanent split capacitor induction motors (SPPSCIM). This paper presents a networked Variable Frequency Drive (VFD) for these ACs to optimize operation, enhance energy efficiency and increase load drive capability of constrained energy systems. The network mode of operation allows existing installed base of standalone units to be integrated with central control infrastructure in an organization through Building Management Systems (BMS). The drive is designed, and tested on a 2.3kW Mitsubishi Electric Split-type air conditioner with SPPSCIM. The results of implemented VFD with networked control over Wi-Fi are presented to validate the effectiveness of the speed drive and the enhanced capability enabled by networked operation.

Networked PFC drive for SPPSCIM air conditioners to enable elastic load operation in constrained power systems

Large HVAC systems are good candidates for elastic loads in demand side management optimization of smarter grid infrastructures. There is an increaseing trend of using Split-type Air Conditioners (ACs) in residential, commercial and industrial setups for fine granularity temperature control. The traditional control in each unit is through duty-cycle based hysteretic method which is simple but inefficient and requires power spikes at every start-up of the compressor. In developing countries with weak grids, people rely on backup sources like generators and UPS. Large power surge at each turn-on inflicts huge pressure on these constrained resources. Limited work has been reported that aims at energy optimization and controlled operation of Split-type ACs constructed using single phase permanent split capacitor induction machine (SPPSCIM). Our work targets operational optimization, efficiency enhancement and increased load handling capability of contrained power systems in setups employing multiple AC units. This is done through networked Variable Frequency Drives (VFD) for conventional ACs which allow elimination of surge current at turn-on and tighter set-point control while drawing sinusoidal current near unity power factor from the constrained source. Furthermore, the networked operation allows integration of existing installed base of standalone units through Building Management System (BMS) with the central control infrastructure. The prototype networked VFD is tested on a 2.3kW Mitsubishi Electric Split-type AC through Wi-Fi and the results are presented to confirm the expected results in network mode of operation.

Smart DC microgrids: Modeling and power flow analysis of a DC Microgrid for off-grid and weak-grid connected communities

This paper presents a modular, scalable and viable architecture for a DC Microgrid to supply electrical power to off-grid communities, particularly in developing countries, where grid electricity is unavailable or is largely intermittent. We have modeled a system for a 40 house village where each house generates (through PV), stores and can share electrical energy with the community for an optimized system. The concept is then extended to an islanded DC Microgrid model with scalable renewable generation and smart power regulation. This paper also includes the operational analysis of the proposed microgrid architecture using Newton Raphson method modified for DC grid analysis. This analysis is used to optimize the voltage level for regulation, efficiency and safety. The results are then verified through an experimental setup in the laboratory. Lastly, this paper also proposes a protection scheme for the DC microgrid and the contributing households.

High efficiency modified dual-active bridge converter for photovoltaic integration

The penetration of distributed renewable energy resources especially photovoltaic (PV) panels is expected to increase over the next few decades. We propose a modified dual-active bridge dc-dc converter for integration of PV panels. The converter uses a novel switching scheme for its boost operation and enables power transfer from a low voltage port to a high voltage port. The converter operates in discontinuous conduction mode (DCM), simplifying control and achieving high efficiency results through soft switching. Analysis of converter along with simulation and experimental results are provided to validate the working of proposed converter design.

Analysis on central and distributed architectures of solar powered DC microgrids

In this work, central and distributed architectures of DC microgrids for rural electrification are analyzed under various operating conditions. In the proposed scheme, a single household consumer forms the atomic nanogrid unit which may integrate its resources in a scalable model with the community to form a microgrid, without dependence of the national grid. The flow of power between houses and the microgrid is implemented through a bidirectional flyback converter. The operation of proposed scheme for two different architectures, i.e. distributed generation distributed storage architecture (DGDSA) and centralized generation centralized storage architecture (CGCSA) is evaluated at various distribution voltage levels and conductor sizes. Modified Newton Raphson Method based analysis is performed for both architectures which show that distributed architecture has significant advantages over central architecture due to higher efficiency, low voltage drop and lower line losses. Further, the scalable nature with minimum installation cost for distributed architecture makes it more favorable for rural electrification applications in comparison to central architecture. The simulated results are also verified using a scaled down version of hardware implementation.