Residual-based Attention Physics-informed Neural Networks for Spatio-Temporal Ageing Assessment of Transformers Operated in Renewable Power Plants

Ibai Ramirez,Joel Pino,David Pardo,Mikel Sanz,Luis del Rio,Alvaro Ortiz,Kateryna Morozovska,Jose I. Aizpurua

2024-10-03

Abstract:Transformers are crucial for reliable and efficient power system operations, particularly in supporting the integration of renewable energy. Effective monitoring of transformer health is critical to maintain grid stability and performance. Thermal insulation ageing is a key transformer failure mode, which is generally tracked by monitoring the hotspot temperature (HST). However, HST measurement is complex, costly, and often estimated from indirect measurements. Existing HST models focus on space-agnostic thermal models, providing worst-case HST estimates. This article introduces a spatio-temporal model for transformer winding temperature and ageing estimation, which leverages physics-based partial differential equations (PDEs) with data-driven Neural Networks (NN) in a Physics Informed Neural Networks (PINNs) configuration to improve prediction accuracy and acquire spatio-temporal resolution. The computational accuracy of the PINN model is improved through the implementation of the Residual-Based Attention (PINN-RBA) scheme that accelerates the PINN model convergence. The PINN-RBA model is benchmarked against self-adaptive attention schemes and classical vanilla PINN configurations. For the first time, PINN based oil temperature predictions are used to estimate spatio-temporal transformer winding temperature values, validated through PDE numerical solution and fiber optic sensor measurements. Furthermore, the spatio-temporal transformer ageing model is inferred, which supports transformer health management decision-making. Results are validated with a distribution transformer operating on a floating photovoltaic power plant.

Machine Learning,Systems and Control

What problem does this paper attempt to address?

The paper attempts to address the issue of aging assessment of transformers in renewable energy power plants. Specifically, the paper focuses on the following points: 1. **Accurate Measurement of Hot Spot Temperature (HST)**: Traditional methods for measuring hot spot temperature are complex and costly, usually relying on indirect measurements. These methods cannot provide accurate spatial and temporal resolution, leading to less precise assessments of transformer health. 2. **Modeling of Spatial and Temporal Temperature Distribution**: Existing thermal models typically ignore spatial distribution and only provide worst-case hot spot temperature estimates. This limits a comprehensive understanding of the transformer's actual operating conditions. 3. **Transformer Aging Assessment**: Current aging assessment methods are mainly based on worst-case hot spot temperature estimates, ignoring spatial and temporal variations under actual operating conditions. This may lead to inaccurate maintenance decisions and resource wastage. To address these issues, the paper proposes a method based on Physics-Informed Neural Networks (PINN), combined with Residual Attention Mechanism (RBA), to improve the spatial and temporal resolution of transformer oil and winding temperatures, and to enable accurate aging assessment of transformers. This method allows for more accurate monitoring of transformer health, supporting more effective maintenance decisions.

Residual-based Attention Physics-informed Neural Networks for Spatio-Temporal Ageing Assessment of Transformers Operated in Renewable Power Plants

Predicting transformer temperature field based on physics‐informed neural networks

Artificial Neural Networks Application for Top Oil Temperature and Loss of Life Prediction in Power Transformers

Real-time Health Monitoring of Heat Exchangers using Hypernetworks and PINNs

A hybrid ensemble learning approach for the aging-dependent reliability prediction of power transformers

A Comprehensive Analysis of PINNs for Power System Transient Stability

Integrating Real-Time Monitoring and Asset Health Prediction for Power Transformer Intelligent Maintenance and Decision Support

Physics-Informed Neural Networks for Time-Domain Simulations: Accuracy, Computational Cost, and Flexibility

Research on Power Transformer Health Indicator Based on Improved Convolutional Neural Network

Remaining useful lifetime estimation for discrete power electronic devices using physics-informed neural network

Residual LSTM-based short duration forecasting of polarization current for effective assessment of transformers insulation

Temporal Modeling for Power Converters With Physics-in-Architecture Recurrent Neural Network

HxPINN: A hypernetwork-based physics-informed neural network for real-time monitoring of an industrial heat exchanger

Transformer hot spot temperature estimation through adaptive neuro fuzzy inference system approach

Application of back propagation neural network in complex diagnostics and forecasting loss of life of cellulose paper insulation in oil-immersed transformers

Physics-Informed Neural Networks for Prognostics and Health Management of Lithium-Ion Batteries

Integrating Physics-Informed Neural Networks into Power System Dynamic Simulations

Top-Oil Temperature Prediction of Power Transformer Based on Long Short-Term Memory Neural Network with Self-Attention Mechanism Optimized by Improved Whale Optimization Algorithm

A Deep Learning Approach to the Transformer Life Prediction Considering Diverse Aging Factors

PINNsFormer: A Transformer-Based Framework For Physics-Informed Neural Networks

An approach to aging assessment of power transformer based on multi-parameters