Modular Permanent Magnet Machine Drive System With Fault Tolerant Capability

Author : Hao Zeng
Publisher :
ISBN 13 :
Total Pages : 0 pages
Book Rating : 4.:/5 (139 download)

DOWNLOAD NOW!

Download or read book Distributed Control and Self-sensing of Permanent-magnet Modular Motor Drives (mmds) for Electric Aircraft Propulsion written by Hao Zeng and published by . This book was released on 2023 with total page 0 pages. Available in PDF, EPUB and Kindle. Book excerpt: Permanent magnet (PM) modular motor drives (MMDs) have been recognized as promising candidate motor drive units for electric aircraft propulsion applications due to their high power density, high efficiency, and high level of fault tolerance. A modular motor drive has a modular stator design that consists of a number of multi-phase modules, with each module excited by its own drive. The drive system reliability can be greatly improved by the power stage redundancy introduced by the modular motor drive configuration. However, conventional centralized control is still a potent source of single-point failures that can seriously degrade the MMD's system reliability, requiring the modularization concept to also be implemented in the controls.Heterarchical distributed control has been identified during this research as the preferred control architecture to meet demanding system reliability requirements by offering an independent controller in each module to eliminate single-point failures. In this architecture, the MMD's distributed module controllers operate as equal peers without critical data communications between them in order to avoid cascaded failures. To facilitate distributed control algorithm design and analysis for this heterarchical control architecture, a module-level machine analysis and modeling framework has been developed. This analysis has identified and characterized the key electromagnetic asymmetries in modular machines including intra-module unbalanced inductances and inter-module unbalanced cross-coupled fluxes that are attributable to asymmetric mutual flux couplings within and between the modules, respectively. A generalized complex vector module-level machine model has been proposed for PM MMDs that captures and highlights the asymmetric features that distinguish the modules from conventional machines. It is shown that this developed complex vector model is also compatible with machines with rotor saliencies. Negative sequence current responses are identified that are induced by module-level machine asymmetries. This feature degrades the inter-module independence of MMDs, especially under fault conditions where maximum inter-module independence is desired. A distributed control algorithm with a space vector resonator-based negative sequence harmonic regulator is proposed to implement control-based virtual electromagnetic isolation. With the proposed algorithm, negative sequence responses are rejected to achieve appealingly high levels of drive module independence. A distributed self-sensing algorithm has also been developed during this research program to eliminate single-point failures attributable to the machine's single shaft-mounted position sensor. A distributed current observer/ back-EMF state filter with integrated bandwidth-partitioned observer command feedforward has been developed with virtual electromagnetic isolation features that reject disturbances from other modules on estimated position and speed. A higher level of inter-module independence is achieved in the modular motor drive using the proposed distributed self-sensing algorithm. MMDs with high power densities and high power ratings typically result in high fundamental excitation frequencies at high speeds and limited PWM switching frequencies that often interact to create performance complications. Cross-couplings caused by the PWM latch delay, control update delay, and rotation can significantly degrade the discrete-time control performance of the MMD at high fundamental frequencies. To address this problem, an enhanced discrete-time control framework with manipulated and disturbance input decoupling has been developed, and a direct-synthesized complex vector current regulator has been proposed with an accurate pole-zero cancellation feature to address issues arising from the aforementioned cross-couplings. The proposed discrete-time algorithm has been demonstrated to significantly enhance the MMD's dynamic performance at low control-to-fundamental frequency ratios with values less than 5:1. Experimental evaluation of key features of the proposed distributed current vector control have been carried out using a prototype 200 kW PM modular motor drive system and a low-power PM modular motor drive concept demonstrator unit. Excellent control independence and control coordination in the heterarchical control architecture have been demonstrated under healthy conditions and with various faults based on the proposed distributed control framework and algorithms. The proposed discrete time form of complex vector current regulators has been tested and successfully verified using prototype 200 kW and 1 MW PM modular motor drive systems. The tests have demonstrated excellent dynamic and steady state performance at low control-to-fundamental frequencies ratios as low as 2.5.