Online Short-Circuit Current Estimation of Permanent Magnet Motors

In this article, online short-circuit estimation of permanent magnet synchronous motors (PMSMs) are presented, which is essential for postfault ac drive operations. For this purpose, initially, the PMSM model with interturn short-circuit fault (ITSC) is extended to cover motors with saliency. By utilizing ITSC model voltage equations and voltage reference equations, the amplitude and phase relationship between short-circuit current and the phase reference voltage is established. In the next step, second harmonics in the reference voltage and feedback current are used to extract the short-circuit current information. In order to efficiently use ac drive microcontroller bandwidth, the proposed short-circuit current estimation method is executed after the ITSC fault is detected, not continuously. Instead of running the proposed short-circuit current estimation continuously, it is reasonable to run it after an ITSC fault detection to efficiently use ac drive microcontroller bandwidth. Therefore, two intermediate steps, a simple ITSC fault detection and severity estimation algorithms are proposed as well. The proposed technique can estimate the fault current and severity without the knowledge of fault location and valid for all PM motors. The findings are experimentally tested using two PMSMs to demonstrate the effectiveness and practicability of the proposed estimation technique.

Automated summary

This article presents an online estimation method for short-circuits in permanent magnet synchronous motors (PMSMs), which is crucial for postfault AC drive operations. The method extends the PMSM model to cover motors with saliency and establishes the relationship between short-circuit current and phase reference voltage. Second harmonics in the reference voltage and feedback current are then used to extract the short-circuit current information. By executing the estimation method after the short-circuit fault is detected, the AC drive microcontroller bandwidth is efficiently utilized. Simple fault detection and severity estimation algorithms are also proposed. The experimental results demonstrate the effectiveness and practicability of the proposed estimation technique.