Power Quality Under Different Operating Conditions Of Power Systems

This work is mainly devoted to investigate exploiting the aptitude of artificial intelligence­based techniques and new signal analysis methods in assessing power quality problems, and mitigating their adverse effects. Work in this study has contributed into some important power quality issues as given below. br A new method based on the wavelet multiresolution signal decomposition for the extraction of an apt power quality disturbance feature vector is proposed. The vector consists of only four elements drastically reducing the computational burden. A neurofuzzy classifier trained with simulated event signals is adopted to diagnose 10 practical disturbance categories. br Modified versions of commonly applied FFT­based power quality indices are suggested, avoiding its unwanted drawbacks via utilising the potential of the wavelet multi­resolution signal decomposition. Moreover, load nonlinearity indicator (LNI) is proposed as a novel power quality index. The new set of indices is found to be more accurate, practical and informative. br A universal power quality index is established. Knowledge acquisition from literature and experts provides the modules of the influencing factors, the significant set of factors in each module, its ranking and the manner they affect. Analytic hierarchy processing and fuzzy reasoning are the assistant tools employed to propose this new global measure. br A three­phase four­wire shunt active power filter based on ANN and sliding mode control (SLMC) has been proposed. ANN is applied to detect the line reference current from the nonlinear load current. SLMC is utilised to track the reference current. Also, another version of the filter fully based on neural networks has been proposed and analysed. br A multipurpose shunt three­phase active power conditioner (APC) is proposed. It compensates for all unwanted components in the load. It can also regulate the load terminal voltage. An extended three­phase form of the three­dimension (3D) power concept is exploited to estimate the reference currents for the APC. ANN is employed for the APC current control. br A power quality manger based on a shunt and a series controlled PWM converters is proposed. Its function is to compensate for periodic and non­periodic waveform distortions. A modified synchronous reference frame (SRF)­ fuzzy logic control scheme is employed for this manager. Two configurations are presented and compared. br A universal conditioner comprising series active filter and shunt passive filters is presented. Its function is to compel the line current to be sinusoidal and in phase with the source voltage. Besides, it can damp voltage flickers, and rapidly regulate the load terminal voltage against source voltage sags and swells. A new current control strategy of the series filter utilising the advantages of fuzzy logic is adopted for the device