This paper proposes the modern approach using Particle Swarm Optimization (PSO) algorithm in determining the ideal value of Proportional Integral (PI) gain for current controller of Permanent Magnet Synchronous Motor (PMSM). Controlling the torque of PMSM and optimizing the PI-gain are the main objectives of this project. The PI controller is employed to control the speed and the torque of the PMSM with the implementation of Field Oriented Control (FOC) method. This new proposed PSO technique proved that the ability in reducing the torque ripple compared to conventional heuristic method. The ideal PI-gain acquired from the PSO was included into current PI controller. From the result obtained, it shows that the viability of the PSO technique is the best to determine PI-gain for current controller.

No abstract

Voltage sag and swell can cause serious problems like instability, short lifetime, and data errors in power quality. The objective of this paper is to present the detection and classification of voltage sag and swell. S-Transform is used as a base to detect the triggering point of disturbances using Root Mean Square (RMS) method. This paper also presents the type of sags and swells by applying the features into Extreme Learning Machine (ELM) neural network approach in MATLAB. In addition, ELM method is compared with Support Vector Machine (SVM) and Decision Tree method to observe the best classification between these three methods. The accuracy of the classifications was displayed in percentages. It was verified that the detection using RMS and classification using ELM are possible because the results are clearly showing the advantages of the RMS in detecting and ELM for classifying the power quality problems.

No abstract