Abstract.In this paper, a maximum power point tracker method using sliding mode control for a photovoltaic system is presented. The system includes a photovoltaic array, a DC/DC converter and a DC/AC inverter connected to a load. The designed control regulates the converter output voltage and it maximizes the power generated by the photovoltaic array. To obtain it, the sliding surface used to control the DC/DC converter is adjusted according to PV array output power. The control law designed and the results in a simulation platform will be presented.

This paper presents a hybrid methodology to analyze electromagnetic transients in a photovoltaic distributed system. The methodology consists in split the system into two parts: the external zone, where the system is reduced by a dynamic equivalent obtained by the application of Balance Realization (BR) theory and the internal zone, where the system is modeled in detail and solved by a time domain technique. The methodology, that uses a predictor-corrector method to join both zones, does not need a transmission line to interconnect the systems but an element with reactive behavior to deal with the time coupling. In the first zone, the use of BR permits to obtain a new and reduced state-space description of the system that keeps the domain dynamics of the full system. The size of this system can be reduced as much as desired. Nevertheless, the resulting size is proportional to the accuracy. The BR decreases significantly the computational cost to simulate distributed networks. No restrictions are done in the internal zone where all the dynamics elements including the control if desired, can be simulated. The internal zone commonly contains non-linear or power electronic elements.

Abstract.The non-quality phenomena of the supply voltage in electrical power systems include transient disturbances as frequency variations, sags, swells, flicker or interruptions. In this work, a method to detect and measure some transient disturbances based on Artificial Neural networks (ANNs) will be presented. A Feedforward network has been trained to detect the initial time, the final time and the magnitude of a voltage disturbance. The design and training process of an ANN specialized in voltage sags detection will be presented. The performance of the designed measure method will be tested in a simulation platform designed in Matlab/Simulink through the analysis of a practical case.

No abstract