Reduced Order LPTN Model for Electrical Machine Thermal Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing thermal models for electrical machines are complex and require significant computational resources, often necessitating long processing times and the presence of temperature sensors close to the estimation point, limiting their effectiveness in real-time monitoring and control.
Innovation Solution
A method utilizing a reduced order lumped parameter thermal network (LPTN) model, which selects nodes for sensitivity analysis or proximity-based simplification, allowing temperature estimation from distant locations with reduced computational resources and improved accuracy through iterative parameter adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a full LPTN model is used for temperature estimation, then measurement precision is improved, but computing time increases significantly
Solution Approach 1:
The LPTN model is divided into two parts: a reduced order model for fast computation and a full model for accuracy-critical nodes. This segmentation allows the system to use computational resources efficiently while maintaining measurement precision where needed.
Solution Approach 2:
Different nodes in the LPTN model are treated differently based on their importance. Critical nodes use the full model for accurate temperature estimation, while less critical nodes use the reduced order model, optimizing the balance between precision and computing time.
2Measurement precision
If temperature sensors are placed close to the estimation point, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The LPTN model acts as an intermediary that translates temperature measurements from accessible locations into accurate temperature estimates for critical nodes. This eliminates the need to place sensors directly at difficult-to-reach estimation points while maintaining measurement precision.
Solution Approach 2:
Instead of placing physical sensors at every critical node, the system uses thermal model copying where the LPTN model replicates the thermal behavior of distant nodes based on measurements from accessible locations, providing accurate temperature estimates without additional sensor complexity.
3Productivity
If a reduced order LPTN model is used, then productivity is improved through faster processing, but measurement precision deteriorates
Solution Approach 1:
The system dynamically selects which model (reduced order or full) to use for each node based on the required measurement precision and available computational resources. This dynamic approach allows the system to optimize productivity while maintaining adequate precision for each specific estimation task.
Data Source
Figure 1~3B
Figure 4A~4B
AI summary
A method of monitoring the thermal behaviour of an electrical machine by means of a lumped parameter thermal network, LPTN, model, the LPTN model being formed of interconnected subnetworks describing different parts or portions of the electrical machine, wherein the method comprises: a) selecting a first node in the LPTN model, b) obtaining a reduced order LPTN model of the LPTN model, based on a sensitivity analysis of the LPTN model with respect to the first node or based on proximity of subnetworks with respect to the first node, the reduced order LPTN model comprising simplified subnetworks of the LPTN model and non-simplified subnetworks of the LPTN model, c) obtaining a temperature measurement from a location of the electrical machine which is represented as a second node in a non-simplified subnetwork or from a position of the electrical machine which is within a predefined maximum distance from a location of the electrical machine which is represented as a second node in a non-simplified subnetwork, and d) estimating the temperature in the first node based on the temperature measurement using the reduced order LPTN model.