Energy Management in More Electric Aircraft through PMSM Fault Diagnosis, Adaptive Load Shedding and Ecient Aircraft Design

Abstract

More electric aircraft is an electrifcation scheme of aircraft system with high technical feasibility and good economy. It can reduce the weight of aircraft structure, improve maintenance effciency and reduce fire hazards. However, the electrifcation of aircraft system will drastically increase the proportion of electrical equipment, the total power demand and the diffculty of fault diagnosis. This paper uses the energy management method to take up the challenge, with
focus on fault diagnosis of permanent-magnet synchronous machines (PMSMs), adaptive load shedding and energy effcient aircraft design.
A literature review of the concept evolution from all/more-electric aircraft to energy-optimized aircraft is presented. The main issues of the aircraft electrication process are summarized, and followed by an introduction to the current research and methods. The model of the aircraft electrical system is qualitatively and mathematically recalled, including the generator, the battery, the DC motor, the AC motor, and the electric power converter. The accuracy and computation cost of the aircraft model depends on the complexity of the subsystem models that are involved. Therefore, the level of detail that is necessary for a good precision-versus-simulation-time ratio is discussed by taking the electric system of an industrial level hybrid energy quadcoptor UAV as an example. The analysis shows that the bi-directional instruments, i.e. the electric machine, should be modeled in details while other
components can be simplifed. PMSMs are a group of on-board electric machines with promising future prospects because of high power density and stability. The model of PMSMs is further developed in this work, especially in the inter-turn and phase-to-phase short-circuit conditions. In case of inter-turn short-circuit fault, a winding-function-based and a fault-current-based model are separately
developed. The accuracy of both models are verified and compared through experimental results. The fault-current-based modeling method is applied to the phase-to-phase short-circuit fault and experimentally examined and discussed.
General condition monitoring methods require the use of a large number of sensors. A
fault detection and isolation method that can have low requirement of sensor is recalled and inherited. The description of the fault phase identication index using this method is relatively imprecise, which is not applicable to the inter-turn short-circuit fault. In this work, the analytical
expression of the faulty phase identification index is derived based on the fault models. A method to isolate inter-turn and phase-to-phase short-circuit faults is proposed by a combination of the current- and the voltage-signature residuals. This development expands the application scope of the original fault detection and isolation tool and improves its accuracy.

Date of Award	3 Jun 2019
Original language	English