Development of contactless electromagnetic sensing technologies for key electric devices in microgrids

Abstract

Microgrids attract more and more attention around the world due to the reduced power loss, load burden relief, improved power quality, decreased carbon dioxide emission level, and the eco-friendly benefit. Permanent magnet synchronous generators (PMSGs), power converters, transmission systems, and other components are key electric devices for the microgrids. Thus, these devices need to be monitored continuously to ensure the microgrids operate in a safe, reliable, and efficient manner. However, the traditional monitoring techniques in these key electric devices are not satisfying in some aspects. These aspects include being invasive to the systems, using too many sensors based on multiple principles, applying complicated computation, high-cost sensors, and recalibration in different conditions or systems. These reasons will result in the complexity and expensive costs of microgrid monitoring systems. In this way, the economic benefits of microgrids will be reduced. This thesis is devoted to developing a number of novel monitoring techniques based on electromagnetic sensing for key electric devices in microgrids, which is capable of overcoming the disadvantages of the traditional monitoring methods. Firstly, the conventional approaches for detecting faults in PMSGs, power converters, and transmission systems are reviewed in Chapter 2. The focus of PMSGs is on detecting the magnet defect and eccentricity faults, while the high-frequency current sensing is the emphasis in the power converter part. Transmission systems concentrate on the power quality, current reconstruction, sag, and tower inclination. Subsequently, a new magnet defect and eccentricity fault detection method using magnetic sensing is proposed in Chapter 3. This method is based on measuring the stray magnetic fields outside a PMSG to monitor the demagnetization and eccentricity faults. The magnetic equivalent circuit (MEC) model is employed to analyze the external magnetic fields when a magnet defect fault or an eccentricity fault exists. Then, corresponding detection indexes are proposed for monitoring these faults in a PMSG. Both finite element method (FEM) simulations and experiments validate the effectiveness of the proposed method for monitoring the magnet defect and eccentricity faults. After that, a new current measurement in high-frequency power converters based on magnetic sensing is presented in Chapter 4. The skin effect model of a rectangular conductor is analyzed. Then, a circle trace is proposed to reduce the skin effect from high-frequency currents, and it is compared with a conventional straight trace in step current response and frequency response analysis. Similarly, the FEM simulation and experiments show a better dynamic response and a larger bandwidth of the circle trace than those of the straight trace in current sensing. Last but not least, Chapter 5 demonstrates feasible condition monitoring techniques in transmission systems with electromagnetic sensing and computational intelligence. The power quality monitoring methodology with artificial neural networks can be applied in different microgrids even with various voltages, currents, and frequencies. Moreover, a new approach based on the artificial immune system is capable of reconstructing currents, detecting transmission line sag, and detecting tower tilt. In the meantime, these methods are verified in different operating conditions.