FPGA implementation of wavelet based techniques for measurement of power quality parameters

Abstract

Rapid growth in high-power semiconductor devices and electronic converters technologies facilitates the efficient and economical utilization of distributed and renewable energy sources in the emerging power system. Local power control equipment, power electronic-based equipment and protection circuits are also being used in domestic and industrial sectors. The high proliferation of these non-linear loads has led to the deterioration of power supply quality from the transmission to distribution level. Power quality monitoring is a procedure of gathering, investigating and interpreting raw electrical measurement data into useful information. The power quality monitoring helps in energy management, preventive maintenance, quality control and the design and development of metering instruments. Therefore, continuous monitoring of power supply is an essential step for improving power quality. In the actual power system, voltage and current signals are dynamic and highly unpredictable. Hence, the time-varying nature of the power signal requires a technique that can analyze all sorts of disturbances accurately. In the last few decades, several techniques have been reported in the literature to analyze time-varying distortions. Still, only a few methods are suitable for real-time estimation of harmonics and power quality indices. Further, these methods need improvements concerning the accuracy, implementation complexity, response time and resource utilization. The primary objective of this thesis is to estimate harmonics and power quality indices in real-time while maintaining reasonable accuracy. Hardware efficient algorithms have been developed in this thesis based on the signal processing techniques to fulfill this objective. The Undecimated Wavelet Packet Transform (UWPT) is a multiresolution technique with additional features of redundancy and time invariance. Further, this technique can be easily implemented on the hardware platforms because of its low implementation complexity. This thesis utilizes the potential of UWPT and customizes it to measure harmonics and power quality indices. The work carried out in this thesis is segregated into four parts. The first two work develops a simple and efficient method based on UWPT for accurate estimation of harmonics and power quality indices of time-varying power signal in only one fundamental cycle. The main emphasis of these works is on designing and developing suitable hardware architectures for the implementation of algorithms on the Xilinx Virtex-6 FPGA ML-605 Board. The results of simulated and practical signals are presented to show the advantages of these methods. In third work, the time-invariant property of the UWPT has been exploited for visualization of electrical disturbances in the power system. Further, a measurement technique based on UWPT and Hilbert transform is proposed to estimate instantaneous power quality indices. The simple nature of this method and capability of parallel processing make it feasible to implement on Virtex-6 FPGA ML-605 Board. Results reveal that this method performs better than similar techniques in the literature.