Automated Multistep Parameter Identification of SPMSMs in Large-Scale Applications Using Cloud Computing Resources.

Area of Science:

• Electrical Engineering
• Machine Learning
• Control Systems

Background:

• Parameter identification for permanent magnet synchronous machines (PMSMs) is a mature field.
• Estimating parameters for multiple PMSMs in large-scale applications remains an underexplored challenge.
• Existing methods often require machine information, signal injection, or extra sensors, increasing complexity and cost.

Purpose of the Study:

• To develop a flexible, automated parameter identification strategy for surface PMSMs (SPMSMs) in large-scale systems.
• To overcome limitations of current parameter identification schemes, minimizing complexity, cost, and human intervention.
• To enable efficient parameter estimation without requiring prior machine information or signal injection.

Main Methods:

• A novel multistep parameter identification approach using measurement data from various operating conditions.
• Implementation via an Internet of Things (IoT)/cloud architecture for massive, automated identification.
• Validation using hardware-in-the-loop (HIL) testing.

Main Results:

• The proposed method successfully identifies SPMSM parameters without signal injection, extra sensors, or machine data.
• The IoT/cloud architecture facilitates large-scale, automated parameter identification.
• HIL results confirm the effectiveness and robustness of the developed strategy.

Conclusions:

• The novel multistep approach provides an effective and automated solution for SPMSM parameter identification in large-scale applications.
• The integrated IoT/cloud architecture enables efficient and scalable implementation.
• This work significantly advances the field by addressing the challenge of multi-machine parameter estimation.