Real time high precision harmonic analysis of signals in power systems using HW/SW Co-Design methodology on FPGAs

Increase in use of nonlinear devices in power systems has caused serious harmonic pollution. There is always a need to measure these harmonics for power quality assessment and in cases for mitigating these harmonics. The problem of precision measurement becomes irksome because of effects of frequency deviation. Over the course of time a whole slew of digital signal processing algorithms have been proposed to accurately analyze harmonics in scenarios of frequency deviation but they come with a much increased cost in computational complexity. These computational requirements make them impractical for use in real-time monitoring systems. A computationally complex but highly precise algorithm has been selected for integration into real-time harmonic analysis systems. Due to its complexity it owes its suitability only in offline harmonic analysis system. We have proposed a scheme to realize this algorithm in a manner so that it meets real-time constraints. We have partitioned the algorithm using a Hardware/Software (HW/SW) Co-Design methodology to optimize performance and resource utilization. The design fits well onto low cost Field Programmable Gate Arrays (FPGAs) which have moderate resource count. The computational time advantage achieved by our scheme is compared with an embedded software implementation and performance gain has been presented.