V. Fernão Pires

Power quality disturbances classification using the 3-D space representation and PCA based neuro-fuzzy approach

In this paper a new approach for power quality (PQ) event detection and classification is proposed. This approach is based on an automatic four step algorithm. First the acquired voltage signals are represented in a 3-D space referential. Then principal component analysis is performed. In the third, features are extracted from the obtained eigenvalues of each disturbance waveforms. Finally a neuro-fuzzy based classifier automatically classifies the PQ disturbances. To show the effectiveness of the proposed method several case studies are presented. From the obtained results it is possible to confirm that the proposed approach can effectively classify different PQ disturbances.

Advanced control methods for power electronics systems

The application of modeling methods suitable for control and simulation of power electronic systems is outlined as a self-contained approach to solve the simulation and control problems of novel structures of power electronics converters. The straightforward non-linear modeling for controller design and simulation uses switched state-space models avoiding the averaging task, needs few linear control concepts, derives the stability study from geometric properties and leads to an integrated design of the control, modulators and simulation tasks. On-line sliding mode control techniques are well suited to power converters as they are inherently variable structure systems. Obtained controllers are robust concerning converter parameter variations, semiconductor non-ideal characteristics, load and line disturbances. Main modeling and design steps are summarized and some examples given. Results show fast dynamics, no steady-state errors and robustness against semiconductor non-idealities and dead times.

The application of neural networks and Clarke-Concordia transformation in fault location on distribution power systems

This paper presents a new approach to fault location on distribution power lines. This approach uses an artificial neural network based learning algorithm and Clarke-Concordia transformation. The /spl alpha/,/spl beta/,0 components of line currents resulting from the Clarke-Concordia transformation are used to detect all types of fault. The neural network is trained to map the nonlinear relationship existing in fault location equations. The proposed approach is able to identify and locate all different types of faults (single line to ground, double line to ground, line-to-line and three-phase short-circuit). This approach is subdivided into several main steps: Data acquisition, corresponding on three-phase current signals; Mathematical treatment by the Clarke-Concordia transformation; Fault identification, obtained by the analysis of fault and pre-fault data; Fault location artificial neural network based learning algorithm. The fault position is presented as the output of the neural network on which, as the input, it was considered the eigenvalue of matrix representing transformed line current. Results are presented which shows the effectiveness of the proposed algorithm for a correct fault location on distribution power system networks.

Development of an experimental system for teaching induction motors with fault detection and diagnosis capabilities

The study of fault detection and diagnosis of electrical machines, in particular induction motors, as well as the teaching of their behavior requires practical experience in this field. To provide this experience to the students this paper presents an experimental system that can be used with a standard industrial electrical induction machine. This system is based on only one healthy induction machine that is operated under faulty operating modes. Several fault types such as stator winding faults, bearing faults, or broken bars can be tested. The developed system also allows testing the behavior of the electrical machine for different load types. Several typical loads in which the torque depends of time or speed were implemented. The system is based on computer equipment with a DSPIC, a testing system and an induction machine.

Stator winding fault diagnosis in induction motors using the dq current trajectory mass center

This paper proposes the use of the park transformation mass center applied to the stator currents as a method for diagnosing the occurrence of stator winding faults in induction motor. Induction motor stator currents are first measured and recorded. Then, the park transform is applied to the obtained currents in order to obtain a specific pattern that allows the identification of the stator winding fault. For a healthy motor, a single point in a dq-plane is obtained. However, for an induction motor with some stator winding fault one obtains a circle, in the dq-plane, which is dependent on the fault severity. Accordingly to this relationship a fault severity is reported. In order to show the applicability of the proposed technique, several simulation and experimental results are presented.

A control structure for a photovoltaic supply system with power compensation characteristics suitable for smart grid topologies

A system controller for a three-phase grid-connected inverter for photovoltaic applications is presented. The active and reactive currents will both be controlled in order to realize active power supply and reactive power compensation. The three-phase inverter will be controlled by a fast and robust sliding mode controller. The active and reactive current references will be calculated in order to maximize the electrical energy produced by PV arrays and at the same time as a reactive power compensator. For the calculation of these references the maximum apparent power of the three-phase inverter is also considered. When the PV system is not working at full power the three-phase inverter can also be working as reactive power compensator. The grid injected reactive power is limited by the maximum apparent power of the inverter. This allows integrating this system into a smart-grid. Numerical simulations obtained in the Simulink environment are presented in order to validate the proposed system.

An Eigenvalue/Eigenvector 3D current reference method for detection and fault diagnosis in a voltage source inverter

In this paper a detection and fault diagnosis for a Voltage Source Inverter is presented. The need to insure a continuous and safety operation for voltage source inverter involves the need for reliable fault diagnosis techniques. The proposed approach uses an automatic three step algorithm. Firstly, the inverter output currents are measured which will give typical patterns that can be used to identify the fault. Secondly, the Eigenvectors/Eigenvalues of the 3D current referential are computed. Finally the proposed algorithm will discern if there is any fault and report the faulty switch. The proposed approach was experimentally implemented and its performance verified on various types of working conditions.

Quasi-Z-Source Inverter With a T-Type Converter in Normal and Failure Mode

This paper presents a three-phase multilevel quasi-Z-source inverter (qZSI) topology operating in normal and fault-tolerant operation mode. This structure is composed by two symmetrical quasi-Z-source networks and a three-phase T-type inverter. Besides the intrinsic advantages of multilevel voltage source inverters, the proposed structure is also characterized by their semiconductor fault tolerance capability. This feature is only obtained through changes on the modulation scheme after the semiconductor fault and does not require additional extra-phase legs or collective switching states. In certain fault types, the reduction of the output power capacity will be compensated by the boost characteristic of the qZSI. The fault-tolerant behavior of the proposed topology is demonstrated by several simulation results of the converter in normal and fault condition. To validate the characteristics of this multilevel qZSI, an experimental prototype was also built to experimentally confirm the results.

Hybrid cascade multilevel inverter using a single DC source for open-end winding induction motors

A new hybrid multilevel inverter using a single DC source for open-end winding induction motor is proposed. This power converter uses one three phase bridge inverter and three single-phase inverters, while using only one DC source. Due to the topology and voltage values, this power converter works as three single-phase multilevel inverters. In order to maximize the inverter output voltage levels there is an asymmetry between the amplitude of the three phase bridge inverter DC source and the single-phase inverter capacitor voltages. To control the three-phase inverter a SPWM modulator is used. Since the three single-phase inverters do not have DC sources an algorithm is required to maintain the capacitor voltage balanced. In this work a sliding mode control based approach is used to control the capacitor voltage of the single-phase inverters. In order to verify the proposed topology and proposed system converter several simulation results are presented.

Fault tolerant operation of active neutral point clamped multilevel converters using a voltage sliding mode controller

The three-level active neutral point clamped converter (ANPC) allows to overcame the problem of unequal loss distribution among semiconductor devices that appears in the classical neutral point clamped converter NPC. Another advantage of this power converter is that it can operate under faulty conditions. In this paper the study of the ANPC using a voltage sliding mode controller for a vector modulator under a semiconductor open fault is presented. The approach allows keeping the voltage of the neutral point balanced under a fault condition. Simulation results are presented in order to verify the effectiveness of the proposed control strategy under the fault condition mode.