Abstract
A motor drive system should be reliable, efficient, and robust under numerous applications, loads, and control schemes. To ensure such characteristics and fully test the range of operation for a motor drive, manufacturers and developers traditionally must have a wide assortment of test bed equipment to recreate various machine load combinations. During motor drive development stages, expensive hardware is exposed to faulting and instability risks prior to completely debugging the product. The associated risks and expenditures increase with the power level. Therefore, this thesis provides a motor drive testing method, deemed the “virtual machine,” (VM) that removes a great deal of the risk and cost associated with motor drive development and validation. The technique used to accomplish the VM exploits the Power Hardware in the Loop (PHIL) concept to replace equipment; particularly a voltage amplifier is used to recreate the terminal characteristics of various machine loading scenarios that a motor drive is conventionally tested against. A unique transformer coupling network is proposed between the amplifier and motor drive to provide decoupling and properly step voltages. By implementing the VM concept on such a transformer coupled PHIL test bed, potential pitfalls and non-linearities due to the transformer can be assessed; thus, providing the field with a new PHIL filter structure and de risking the design prior any future increments in power level. When validating, the VM shows matching, consistent results compared against both simulations and a physical induction machine (IM) energized via the same motor drive. Note that multiple counter torque loads were provided via a DC machine. Although the proposed amplifier control method is based on a steady state phasor system, it also proves adequate for recreating the transient terminal characteristics of an IM across the line start. The conclusion proves the VM concept a viable solution for removing cost and risk in drive development as well as verifies the PHIL transformer coupling network. The concept of controlling active and reactive power flow between parallel connected LCL coupled converters is established then applied as a PHIL technique; thus, opening the field to use this approach for evaluating more complex, simulated systems. The limitations of the proposed method are discussed as well as future work areas to address such constrictions and improve the fidelity of the VM. Finally, after polishing the limitations, a future direction of increasing the VM power level is established and some derivatives of the PHIL load emulation concept given.
|
