Position Estimation Method of IPMSM in Full Speed Range by Simplified Quadratic Optimization

Abstract

In this paper, a new full-speed-range position estimation method for the interior permanent magnet synchronous machine (IPMSM) is proposed. This method is designed for high power traction motor drives. In such application, variable speed operation from zero to high speed is desired. The objective of this paper is to obtain smooth position estimation in full speed range to guarantee high reliability of the motor control system. Firstly, a quadratic cost function based on the stator voltage equations is constructed. The rotor position in the full speed range is obtained by numerically solving an optimization problem. After obtaining the rotor position, the motor speed is then obtained through a phase-locked loop observer. In such a way, only one variable in the optimization problem has to be solved. Therefore, the numeric calculation burden is greatly reduced. The convexity of the quadratic cost function considering sampling noise as well as parameter mismatch is discussed in detail. Thus, the solvability and accuracy of the estimation result is guaranteed. At last, experiments on a traction motor are conducted to verify the effectiveness of the proposed position estimation method. INDEX TERMS IPMSM, Position sensorless control, Newton method, Full speed range.

Simplified Quadratic Optimization-Based IPMSM Full-Speed Range Rotor Position Estimation in Synchronous Rotating Frame

“…1. The PLL observer used in this article is the same as that used in [29], from where more details could be found.…”

“…However, to further clarify the derivation and simplify the solving process, the optimization problem is formulated based on the IPMSM voltage equations in the synchronous rotating frame. Moreover, after applying the Taylor expansion on the voltage equation in the synchronous rotating frame, the convexity of the problem, the sampling noise, and parameter sensitivity are more straightforward to analyze, and the solving method is also remarkably simplified compared to the similar method presented in [23] and [27]- [29]. In the end, both the effectiveness and the robustness of the proposed method are validated on a traction motor.…”

“…Unified position estimation methods have been proposed to tackle the aforementioned method switching problem for IPMSM control in the full-speed range [23], [25]- [29]. These methods utilize a unified voltage model [23], [27], [28] or flux model [25], [26] in the full-speed range.…”

“…Based on the previous work presented in [28] and [29], this article proposes an improved position estimation method for IPMSM using simplified quadratic optimization in the full-speed range. The proposed method also interprets the position estimation as a single-variable optimization problem.…”

Regression Model-Based Flux Observer for IPMSM Sensorless Control with Wide Speed Range

“…Recently, a unified model-based sensorless method was proposed as a new approach. By allowing additional signal injection, the performance of the model-based sensorless method was significantly improved [20,21]. Applying Newton's method to the nonlinear optimization problem, these methods directly estimated the rotor position angle and speed.…”