Adaptive neuro fuzzy inference system based power quality assessment

Abstract

Power quality improvement has a positive impact on sustained profitability of the distribution utility on the one hand and customer satisfaction on the other. The Total Demand Distortion was introduced in IEEE Std. 519-1992 to measure the current distortion level instead of the total harmonic distortion that was introduced in the earlier version of the IEEE Std. 519-1981. Although the value of the total demand distortion index can quantify the current harmonic distortion level, it cannot reveal whether this distortion level is within or outside permissible limits. Therefore there is a need for an index that can evaluate the current harmonic distortion and reveal whether this distortion is within or outside the allowable limits. An Adaptive Neuro-Fuzzy Inference System based Total Demand Distortion Factor (ANFIS-TDDF) is proposed in this thesis work to evaluate the current harmonic distortion. From the results obtained when applying the proposed ANFIS method into different distortion cases in sinusoidal and nonsinusoidal situations, it is found that the ANFIS-TDDF is able to convey two important information: it can give an indication on how much free is the current waveform from distortion or how much near the sinusoidal wave shape and also decide whether the distortion contained in the current is within the acceptable limit or not, which cannot be obtained from the TDD alone. In this thesis work, an ANFIS based Representative Quality Power Factor (RQPF) is introduced. The proposed ANFIS based Power Quality Assessment methods will be very useful for billing purposes since it has proven to be very sensitive and suitable for all operating conditions. Therefore, customers will be charged the correct penalty, under all conditions. The ANFIS based methods also have the advantages of being simple, easy to be implemented, flexible, easily altered, adjusted and contains its knowledge base. So there is no need for an expert after the designing stage. newline