Temperature prediction of the electric propulsion unit is one of the crucial components to ensure the performance as well as early signature of the faults. However, in the vehicle, the propulsion motor, proper temperature measurement required several sensors and corresponding data post-processing. In the prototype phase, the number of sensors and post-processing is a standard process. However, the motor used in the vehicle has only 1 or 2 sensors. Therefore, it is a challenge to predict the temperature in different locations of the motor. Based on the simulation results using 3D CFD and measured data in the prototype phase, a simplified 1D model is used to predict the temperature of the motor. However, the simplified model cannot predict the temperature properly if dynamically.

The main objective of this work is to predict the temperature of the battery electric vehicle (BEV) propulsion motor using artificial intelligence (AI) that can overcome the shortcomings of the simplified static model. The AI-based thermal model that can predict the temperature more accurately, and it is able to provide fault predicting capability. 

Tasks:

 ·  Data Collection & Preprocessing

• Gather temperature data from prototype phase sensors.
• Integrate simulation results from 3D CFD models.
• Clean, normalize, and preprocess data for AI training.

·  Feature Engineering & Selection

• Identify key thermal parameters affecting motor temperature.
• Extract relevant features from sensor data and simulation results.
• Optimize feature set for AI model accuracy.

·  Development of AI-Based Thermal Model

• Design and train machine learning models (e.g., Neural Networks, XGBoost, LSTMs) for temperature prediction.
• Compare model performance against simplified 1D static models.
• Implement real-time adaptability for dynamic temperature variations.

·  Model Validation & Testing

• Validate AI model predictions using real-world test data.
• Conduct error analysis and fine-tune hyperparameters.
• Ensure reliability in different operational scenarios.

·  Fault Prediction & Anomaly Detection

• Develop AI-based fault detection for early anomaly identification.
• Implement predictive maintenance alerts based on thermal trends.
• Reduce the risk of motor overheating and performance degradation.

·  Performance Evaluation & Continuous Improvement

• Compare AI model predictions with real-world motor temperature readings.
• Enhance model accuracy with incremental learning techniques.
• Conduct long-term testing for durability and robustness.

AI & Machine Learning Expertise

• Deep Learning & ML Techniques: Experience with neural networks (e.g., LSTMs, CNNs) and ensemble models (e.g., XGBoost, Random Forest) for predictive modeling.
• Feature Engineering: Ability to extract and preprocess relevant features from sensor data and simulations.
• Model Validation & Optimization: Techniques for hyperparameter tuning, cross-validation, and performance evaluation.
• Anomaly Detection: Implementing AI-driven fault prediction methods to detect early signs of failure.

Software & Tools

• Programming Languages: Python for model development.
• Machine Learning Frameworks: TensorFlow, PyTorch, Scikit-learn for training AI models.
• Data Processing Tools: Pandas, NumPy, and SciPy for data handling and analysis.

Domain Knowledge & Expertise (Good to have)

• Thermal Engineering: Understanding heat transfer mechanisms, thermal modeling,
• Sensor Technology: Expertise in selecting and calibrating temperature sensors for reliable data acquisition.

This topic is a part of the project called

HeatTrack: Enhanced Reliability, Monitoring and Diagnostics of Complex Cooling Systems through Advanced Thermal Management.

https://www.es.mdu.se/projects/610-HeatTrack__Enhanced_Reliability__Monitoring_and_Diagnostics_of_Complex_Cooling_Systems_through_Advanced_Thermal_Management_

SUSTAINABE ELECTRIC DRIVE