Search

This paper reports on a machine design tool for large (5-10MW) direct drive electrical generator. The aim of the work is to construct a flexible calculation tool that enables the analysis of different ideas and concepts for generator design. The tool is intended for engineers that are involved in the design of wind turbine systems.The design tool comprises calculation modules that are kept as independent as possible from each other so that new machine geometries and types can be modelled by reusing, recombining and modifying the different modules.Choice of the most suitable candidates for the application is presented. Assessment is made based on suitability criteria combined with the SWOT analysis method.An analytic design is generated at first; then visualised in 3D CAD, evaluated using FEM and a dynamic model simulation. An optimisation module is available for improving the design.The calculations were validated by comparing results with literature reporting machines with similar ratings, laboratory tests on a prototype and by evaluating a given design using different software like finite element, 3D CAD and dynamic simulation programs.

This paper presents a 1 kW generator and converter applied to a floating Darreius type wind turbine system. The completed system is floated in the sea with the generator 5m below the surface.The present paper sketches the design and presents the laboratory bench test results, as well as the environment where the turbine was placed.

Nowadays the requirements imposed by the industry and economy ask for better quality and performance while the price must be maintained in the same range. To achieve this goal optimization must be introduced in the design process. Two of the best known optimization algorithms for machine design, genetic algorithm and particle swarm are shortly presented in this paper. These two algorithms are tested to determine their performance on five different benchmark test functions. The algorithms are tested based on three requirements: precision of the result, number of iterations and calculation time.Both algorithms are also tested on an analytical design process of a Transverse Flux Permanent Magnet Generator to observe their performances in an electrical machine design application.

Cogging torque results due interaction of magnetic field of magnets and stator slots, and have negative effects on permanent magnet machines such as vibrations, noise, torque ripples and problems during turbine start-up and cut-in. In order to reduce cogging torque this paper presents a study of influence of closed stator slots on cogging torque using magnetic slot wedges.

The permanent magnet applications based on carbon steel magnets, hard ferrites and AlNiCo magnets classes are renewed with new classes of advanced magnetic materials based on rare earth elements, the Sm-Co and NdFeB types. Performance increase of the hard magnetic materials and their use in specific applications impose also great advances in the field of magnetic measurement. New researches need to be validated in order to investigate the NdFeB permanent magnets performances, including their stability under different thermal operational regimes. In this paper a specific investigation procedure of magnetic performances of NdFeB permanent magnets in correlation with the range of operating temperature is proposed based on modern hysteresisgraph method and impulse magnetization technique.

Comparison of two approaches for the simulation of the dynamic behaviour of a permanent magnet linear actuator is presented. These are full coupled model, where the electromagnetic field, electric circuit and mechanical motion problems are solved simultaneously, and decoupled model, where first a set of static magnetic filed analysis is carried out and then the electric circuit and mechanical motion equations are solved employing bi-cubic spline approximations of the field analysis results. The results show that the proposed decoupled model is of satisfactory accuracy and gives more flexibility when the actuator response is required to be estimated for different external conditions, e.g. external circuit parameters or mechanical loads. Comparison of two approaches for the simulation of the dynamic behaviour of a permanent magnet linear actuator is presented. These are full coupled model, where the electromagnetic field, electric circuit and mechanical motion problems are solved simultaneously, and decoupled model, where first a set of static magnetic filed analysis is carried out and then the electric circuit and mechanical motion equations are solved employing bi-cubic spline approximations of the field analysis results. The results show that the proposed decoupled model is of satisfactory accuracy and gives more flexibility when the actuator response is required to be estimated for different external conditions, e.g. external circuit parameters or mechanical loads.

This paper proposes an optimized version of a new absolute linear encoder (ALE). The innovative ALE can be used for long distance applications (more then 150m) and the accuracy of the measurements is 0.5mm. To obtain these performances the ALE uses a new coding algorithm. This new coding algorithm is the core of the ALE and it allows an economical device solution. The optimized version is able to measure a double distance (more then 300m) with a better accuracy (0.25mm). These performances are obtained using the same device, the same number of sensors and the same ALE structure. The only changes were made in the coding algorithm, in the ruler topology and in the dedicated software. The optimized ALE is a robust device able to work in industrial environment, with a high level of vibrations. By this reason it is ideal for the transport system control in automating manufacturing processes, intelligent storage spaces, huge archives and libraries.

The paper proposes a new solar position sensor used in tracking system control. The main advantages of the new solution are the robustness and the economical aspect. Positioning accuracy of the tracking system that uses the new sensor is better than 1°. The new sensor uses the ancient principle of the solar clock. The sensitive elements are eight ordinary photo-resistors. It is important to note that all the sensors are not selected simultaneously. It is not necessary for sensor operating characteristics to be quasi-identical because the sensor principle is based on extreme operating duty measurement (bright or dark). In addition, the proposed solar sensor significantly simplifies the operation of the tracking control device.

Accurate knowledge of the high frequency inductance profile plays an important role in many designs of sensorless controllers for Surface inductance. A special algorithm is used to decouple the cross-coupling effects between the d-axis and the q-axis, which allows Mounted Permanent Magnet (SMPM) synchronous motors. This paper presents an AC+DC measurement method for determination of the d-axis and q-axis high frequency inductance profiles of SMPM synchronous motors. This method uses DC currents to set a desired magnetic working point on the motor laminations, and then superimpose balanced small AC signals to measure the incremental a separate determination of the d, q inductance profiles as functions of the d, q currents. Experimental results on a commercial SMPM motor using the proposed method are presented in this paper. Accurate knowledge of the high frequency inductance profile plays an important role in many designs of sensorless controllers for Surface inductance. A special algorithm is used to decouple the cross-coupling effects between the d-axis and the q-axis, which allows Mounted Permanent Magnet (SMPM) synchronous motors. This paper presents an AC+DC measurement method for determination of the d-axis and q-axis high frequency inductance profiles of SMPM synchronous motors. This method uses DC currents to set a desired magnetic working point on the motor laminations, and then superimpose balanced small AC signals to measure the incremental a separate determination of the d, q inductance profiles as functions of the d, q currents. Experimental results on a commercial SMPM motor using the proposed method are presented in this paper.

A comparative study of permanent magnet linear actuators with different location of the permanent magnet is reported. Three mover types are considered - soft magnetic mover, permanent magnet mover and hybrid mover. Force-stroke characteristics are obtained with the help of finite element models verified experimentally. Actuator performance was compared on the basis of force per unit power and force per unit current.

The data-based machine learning is an important aspect of modern intelligent technology, while statistical learning theory (SLT) is a new tool that studies the machine learning methods in the case of a small number of samples. As a common learning method, support vector machine (SVM) is derived from the SLT. Here we were done some analogical experiments of the rotor broken bar faults of induction motors used, analyzed the signals of the sample currents with Fourier transform, and constructed the spectrum characteristics from low frequency to high frequency used as learning sample vectors for the SVM. After a SVM is trained with learning sample vectors, so each kind of the rotor broken bar faults of induction motors can be classified. Finally the retest is demonstrated, which proves that the SVM really has preferable ability of classification. In this paper we tried applying the SVM to diagnose the faults of induction motors, and the results suggested that the SVM could yet be regarded as a new method in the fault diagnosis.