Switched Reluctance Motor Torque Ripple Reduction via Offline TSF
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Switched reluctance motor (SRM) drives face high commutation torque ripple due to poor phase current tracking precision, nonlinear inductance profiles, and torque-current-rotor position characteristics, limiting their torque-ripple-free speed range and efficiency.
Innovation Solution
A new family of offline torque sharing functions (TSFs) is developed, using a Tikhonov factor-based objective function to minimize phase current squares and derivatives of current references, which are determined using Lagrange multipliers to reduce torque ripple and copper loss over a wide speed range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional torque sharing functions are used in SRM drives, then the control structure remains simple, but torque ripple increases due to poor phase current tracking precision and nonlinear characteristics
Solution Approach 1:
The patent applies parameter changes by modifying the torque sharing function parameters (α and β coefficients) to optimize the balance between simplicity and performance. The improved TSF uses adjusted parameters to better track the nonlinear torque-current-rotor position characteristics, reducing torque ripple while maintaining the simple control structure of conventional systems.
Solution Approach 2:
The patent implements dynamics by transitioning from static torque distribution to dynamic torque sharing that adapts to varying rotor positions and operating conditions. The improved TSF dynamically adjusts torque distribution among phases based on real-time rotor position feedback, enabling the system to compensate for nonlinear inductance profiles and minimize torque ripple across different operating points.
2Reliability
If offline torque sharing functions are used to reduce torque ripple, then the torque-ripple-free speed range extends, but the computational complexity increases due to Tikhonov factor optimization and Lagrange multipliers
Solution Approach 1:
The patent applies preliminary action by pre-calculating the optimal torque sharing function parameters offline using Tikhonov factor-based objective functions and Lagrange multipliers. These pre-computed parameters are then stored and directly applied during online operation, eliminating the need for real-time complex optimization calculations while maintaining the extended torque-ripple-free speed range achieved through rigorous offline analysis.
Solution Approach 2:
The patent substitutes complex real-time computational mechanics with pre-computed mathematical models. The offline optimization using Tikhonov regularization and Lagrange multipliers replaces the need for complex online calculations, transforming the problem from a dynamic computational challenge to a static parameter lookup operation that maintains reliability without computational burden during operation.
3Object-generated harmful factors
If phase current tracking precision is improved to reduce torque ripple, then torque smoothness increases, but copper loss increases due to higher current derivatives
Solution Approach 1:
The patent applies parameter changes by optimizing the torque sharing function parameters (α and β) to achieve the right balance between current tracking precision and derivative magnitude. The improved parameters enable sufficient torque smoothness while controlling the rate of change of current, thereby reducing copper losses that would otherwise result from aggressive current tracking.
Solution Approach 2:
The patent converts the potentially harmful effect of current derivatives into a benefit by using Tikhonov regularization in the offline optimization. This approach intentionally controls the derivative magnitude to prevent excessive copper loss while still achieving torque ripple reduction, transforming what could be a harmful trade-off into an optimized balance where both objectives are partially satisfied.
Data Source
AI summary
A method for controlling a switched reluctance motor, the method comprising: receiving a reference torque Te ref; receiving an indication of a present rotor position θ for the switched reluctance motor; determining at least one of: a reference current ie_ref(k−1) for a (k−1)th phase, a reference current ie_ref(k) for a (k)th phase, and a reference current ie_ref(k+1) for a (k+1)th phase; and outputting the determined at least one reference current to a current controller operatively coupled to the switched reluctance motor, wherein the determined at least one reference current is based on an objective function comprising the squares of phase current and derivatives of current reference.