Development of a new tunable/adaptive line enhancer for power system harmonic signal retrieval

This thesis deals with Infinite Impulse Response (IIR) adaptive line enhancement and its application to the problem of power system harmonic signal retrieval. A new line enhancer structure is proposed and its parametisation is presented in Chapter 2. The new line enhancer is composed of two filter blocks. One of these is a phase matching filter while the other is a notch filter. The resulting line enhancer is optimal in the sense that it can cancel the broad-band noise entirely if the bandwidth is small enough. The characteristics of the proposed line enhancer are investigated and a simple design approach is given. The simulation results clearly show the performance achieved. The issue of implementing the new line enhancer as an adaptive system is investigated in Chapter 3. Two adaptive line enhancer model structures are proposed and analysed. Error surface analysis is conducted and the results exhibit the convergence characteristics for the different model structures. The application of adaptive techniques such as: the gradient based algorithm, the recursive maximum likelihood algorithm and the approximate maximum likelihood algorithm (proposed by previous researchers) are considered together with the proposed line enhancer. Simulation results are presented and evaluated for the proposed adaptive line enhancement model structures and adaptive algorithms.