INVESTIGATIONS ON OPERATION AND CONTROL OF SELF-EXCITED INDUCTION GENERATOR

Abstract

Growing concerns of carbon dioxide emission and environmental pollution motivated the mankind worldwide towards the utilization of green and clean energy. The various energy sources in the green energy domain are wind, solar, biomass, tidal, ocean waves etc. There is a lot of potential to tap the green energy utilizing for medium and low power applications at isolated places of the globe. Due to vast availability of medium and small generators such as self-excited induction generator, brushless DC generator, permanent magnet generators and switched reluctance generator etc., the possibility of power generation from said energy sources enhanced manifold. The basic problem associated with these energy sources are their intermittent power supply nature. However, the matured power electronics technology has provided the viable solution if its proper interface is provided with the main energy source and load. The self-excited induction generator (SEIG) is the natural choice over other generators for three phase medium power applications due to its better electrical and mechanical characteristics. However, the problem of voltage and frequency variation is the concern for researchers at variable loading even at constant power supply from the prime mover turbine. Moreover, due to extensive use of nonlinear circuits in load, power quality problems such as unacceptable level of harmonics in the supply current, load unbalance at the generator terminal, are an additional issue to address. Further, the power electronics devices like active power filters (VSC/ DSTATCOM) are useful to mitigate the aforementioned problems. The operation of these devices is greatly dependent on the feedback controllers working in the current and voltage loops. There are many pulse width modulation techniques available for generating the gate pulses. The controllers in loops are equipped with control algorithms which are getting various feedback electrical signals at its Analog to digital converter (ADC). The selection of v | P a g e these control algorithms is dependent on their accuracy, speed and capability to extract fundamental components under dynamic loading conditions. Therefore, this thesis deals with the implementation of a few adaptive control algorithms based on least mean square principle, for improving the performance of Self Excited Induction Generator. The SEIG under study is working in standalone mode to feed power at isolated places and connected to linear/nonlinear load. The major problems associated with such kind of standalone power system is regulation of its voltage and frequency under variable loads. Further, maintaining a purely sinusoidal supply current with improved power factor is also one of the major objectives towards power quality standards. A static compensator (STATCOM) is used along with load to compensate currents to maintain power quality of standalone generator under various operating conditions. The Leaky-Momentum Control Algorithm (LMA) is being used to generate switching signals. A voltage source converter (VSC) to enhance the performance of a three-phase self excited induction generator operating with varying loads is used. This LMA technique controls VSC to regulate voltage and frequency of SEIG within a permissible limit. The LMA control is implemented to reduce the higher demand of reactive power, eliminations of harmonics in source current and balancing of loads under different operating conditions. During the electrical and mechanical dynamical conditions, the LMA technique is maintaining a constant voltage and frequency at point of common interfacing (PCI). The proposed technique is a modified control technique of basic Leaky and Momentum Algorithms. This control has removed the drawbacks of Leaky and momentum algorithms. Moreover, it is observed that LMA performs better when there are uncertainties in input conditions. In the second case, a modified NLMS based control algorithm is designed to control the SEIG system feeding three-phase nonlinear loads. The control algorithm extracts the fundamental weight components with reduced oscillations from sensed load current. vi | P a g e Implementation of modified algorithm for control of voltage source converter (VSC) provides fast dynamic response, harmonic eliminations, active/reactive power compensation, load leveling, rapid convergence and enhances the power quality in an isolated distributed wind energy generation system under nonlinear load conditions. The battery energy storage (BES) is employed at DC Link of STATCOM to balance the power in the system during load perturbations and wind speed changes. The simulation study on the proposed system shows the improved steady state and dynamic performance of SEIG system under fixed/varying wind speed while feeding fixed/varying nonlinear load. The entire system comprising SEIG, nonlinear load, voltage source converter and battery storage system is implemented in MATLAB /SIMULINK. Also, the experimental validation of proposed control approach is carried out in the laboratory environment to study the effectiveness. It has shown promising performance under both dynamical state and steady state of the system.