A.J. Pires

Unsupervised Neural-Network-Based Algorithm for an On-Line Diagnosis of Three-Phase Induction Motor Stator Fault

In this paper, an automatic algorithm based an unsupervised neural network for an on-line diagnostics of three-phase induction motor stator fault is presented. This algorithm uses the alfa-beta stator currents as input variables. Then, a fully automatic unsupervised method is applied in which a Hebbian-based unsupervised neural network is used to extract the principal components of the stator current data. These main directions are used to decide where the fault occurs and a relationship between the current components is calculated to verify the severity of the fault. One of the characteristics of this method, given its unsupervised nature, is that it does not need a prior identification of the system. The proposed methodology has been experimentally tested on a 1kW induction motor. The obtained experimental results show the effectiveness of the proposed method

Obtaining the magnetic characteristics of an 8/6 switched reluctance machine: from FEM analysis to the experimental tests

This paper describes the procedure for modeling an 8/6 switched reluctance machine (SRM) step-by-step. First, the geometry and mechanical parameters of the machine are established, when a finite-element modeling (FEM) approach is initially developed to obtain the flux linkage/current/rotor position relationship so that the proper and mutual inductances and also its torque and field distribution characteristics can then be computed. Afterward, a series of experimental tests are performed to obtain the magnetic characteristics of the machine, comparing, correcting, and discussing the results with those of FEM analysis.

Language identification of controlled systems: modeling, control, and anomaly detection

Formal language techniques have been used in the past to study autonomous dynamical systems. However, for controlled systems, new features are needed to distinguish between information generated by the system and input control. We show how the modeling framework for controlled dynamical systems leads naturally to a formulation in terms of context-dependent grammars. A learning algorithm is proposed for online generation of the grammar productions, this formulation then being used for modeling, control and anomaly detection. Practical applications are described for electromechanical drives. Grammatical interpolation techniques yield accurate results, and the pattern detection capabilities of the language-based formulation makes it a promising technique for the early detection of anomalies or faulty behavior.

PCA-Based On-Line Diagnosis of Induction Motor Stator Fault Feed by PWM Inverter

In this paper an automatic PCA (principal component analysis) based algorithm is presented for an on-line diagnostics of three-phase PWM feed induction motor stator fault. The analysis of the fault is derived from the two first principal components of the power inverter output current alphabeta-vector patterns. The obtained eigenvalues are used to discern if the motor is healthy or not, and the correspondent eigenvectors can infer the phase in which the fault occurs. The method is simple to implement and is able to indicate the extend of the fault; rather then only detect its presence

Fault detection using immune-based systems and formal language algorithms

This paper describes two approaches for fault detection: an immune-based mechanism and a formal language algorithm. The first one is based on the feature of immune systems in distinguish any foreign cell from the bodys own cell. The formal language approach assumes the system as a linguistic source capable of generating a certain language, characterised by a grammar. Each algorithm has particular characteristics, which are analysed in the paper, namely in what cases they can be used with advantage. To test their practicality, both approaches were applied on the problem of fault detection in an induction motor.

Entropy-Based Choice of a Neural Network Drive Model

The design of a neural network requires, among other things, a proper choice of input variables, avoiding over fitting and an unnecessarily complex input vector. This may be achieved by trying to reduce the arbitrariness in the choice of the input layer. This paper discusses the relation between the memory range of a particular controlled dynamical system (induction drive) and the dimension of the neural network input vector. Mathematical techniques of process-reconstruction of the underlying process, using coding and block entropies to characterize the measure and memory range were applied. These modeling techniques provide a precise knowledge of the drive dynamics, a fundamental requirement in modern control approaches

Modelling Electromechanical Drive Systems: A Formal Language Approach

Abstract The aim of this paper is to introduce a formal language approach to the modelling of electromechanical drives. Modem applications often involve electromechanical drives with high dynamical complexity. The integration of these systems with information processing techniques allows for new perspectives of modelling their behaviour. The formalism of a formal language-based modelling algorithm is presented and a learning algorithm is proposed for on-line generation of the grammar productions. Grammatical interpolation techniques are also presented to establish new productions by a structural matching procedure. Experimental results are presented and discussed.

Position control of an 8/6 switched reluctance machine without current sensor

The electrical motion drives are more and more used in various industrial processes. The choice of the motor drive is based on power, speed, precision, and cost requirements. There are many different types of motors for motion drives application. AC and DC motors feature brushed or brushless designs are suitable for a variety of industrial applications. Speed, torque and position control are the most used motion control systems. There are many papers about switched reluctance motor speed control and torque control but few about position control. This paper pretends to contribute for the potentiality of the switched reluctance motor in position application. In the first part it is presented the theoretical concepts for motor shaft position. The second part presents a prototype implementation and experimental results when intermediate position between two poles position alignment are reached.

Magnetic Characteristics Model for an 8/6 Switched Reluctance Machine: Analytical Function Approach from Experimental Tests

This paper presents an analytical function approach to determine the flux-linkage characteristics for an 8/6 switched reluctance motor. The design of the proposed model is based on experimental magnetization curves. In order to achieve less expense and time, regarding the simulation and implementation proposes, the model curves are calculated from main parameters normally available or easily measured. Simulation is used to evaluate the model operation. The developed model was experimentally verified and results are presented in this paper.

Study and Analysis of the Switched Reluctance Machine Shaft Position

The switched reluctance machine (SRM) was, during the last decade, target of the attention of researchers. Some studies stand out the application potentialities of the machine, anothers developed the machine design for better performance in terms of efficiency, good torque and speed behavior. Others denote the possibility of good machine functioning without the position sensor and there are still others presenting a way to reduce the noise produced in dynamic regimen. However, there are very few works about the analysis of the machine behavior regarding its shaft position. This work presents some results in this thematic, studying and analyzing its geometry and torque values necessary to develop when a shaft position is intended. Some methodologies are presented in order to be applied to obtain a shaft position. One of them is based on experimental results taking into account the two phases that are necessary to excite simultaneously in order to produce the immobilization of the motor shaft in one determined position.