Page 198 - Decision Making Applications in Modern Power Systems

Adaptive estimation and tracking of power quality disturbances Chapter | 6 161

where :U: denotes the ‘ 1 norm of coefficient vector, and δ is the weight
1
assigned to the penalty term. The cost function is convex, and it is expected
that the EKF algorithm converges to optimum value under some constraints.
The new state update equation for EKF can be expressed as

xðn 1 1Þ 5 xðnÞ 2 ρsgnðxðnÞÞ 1 keðnÞ ð6:37Þ

6.2.4 FPGA implementation of adaptive filters used in power
quality estimation
Adaptive PQ estimation models can be designed by MATLAB/SIMULINK
by implementing the mathematical equations using suitable blocks, which is
shown in Fig. 6.1, but real-time hardware implementation of adaptive filter-
ing based PQ estimation models is quite difficult due to computational
complexity of the model and algorithm. Generally computational complexity
and quantization effects degrade the tracking and estimation accuracy of the
algorithms. Adaptive filtering based PQ estimation model can also be
designed through Xilinx blockset available in MATLAB/SIMULINK library,
which is quite suitable for field programmable gate array (FPGA) hardware
platform. ML506 is an example of general purpose evaluation and develop-
ment platform, and System Generator for DSP is the industry’s leading high-
level tool for designing high-performance DSP systems. Fig. 6.2 shows the
connection of Virtex 5 series board to the laptop. Fig. 6.3 shows the different
sections of adaptive filtering based estimation model designed using Xilinx
blockset.

FIGURE 6.1 SIMULINK modeling of adaptive filter.