New Energy Vehicles

specifically targeting the detection and classification of faults in the drivetrain system. It serves as a critical resource for advancing research and development in electric vehicle safety and reliability. The dataset includes comprehensive data gathered from multiple sensors integrated into the vehicle's operational systems, covering parameters such as voltage, current, motor speed, temperature, vibration, ambient temperature, and humidity. These parameters are essential for monitoring the condition of various drivetrain components and for identifying deviations that signal potential faults.

This dataset encompasses two primary categories of data:

Normal Operational Data: Reflecting the typical performance of a fully functional drivetrain system.
Fault Data: Highlighting malfunctions or inefficiencies in key components, including motors, inverters, and batteries.
The dataset is meticulously curated to ensure a balanced representation of classes, which is crucial for training machine learning models effectively. Each data point is labeled to indicate the system's state, with labels as follows:

Label 0: Normal operation.
Label 1: Motor-related faults.
Label 2: Inverter-related faults.
Label 3: Battery-related faults.
The dataset is intended for use in training and validating advanced deep learning models. These models aim to enhance fault detection capabilities, enabling real-time identification and resolution of issues within NEV systems. By leveraging this dataset, researchers and developers can create more robust diagnostic tools that contribute to improved safety, performance, and reliability of electric vehicles.

Applications of the Dataset
Deep Learning Development: Building convolutional neural networks (CNNs), recurrent neural networks (RNNs), or hybrid models for fault detection.
Predictive Maintenance: Enabling predictive analytics to anticipate potential failures and schedule maintenance proactively.
System Optimization: Enhancing drivetrain efficiency by identifying performance bottlenecks and inefficiencies.
Key Features
High-Quality Sensor Data: Covers a broad range of operational and environmental parameters.
Fault Categorization: Specific identification of faults in critical drivetrain components.
Balanced Dataset: Ensures even distribution of normal and fault data for unbiased model training.
Scalability: Suitable for training models that scale to real-world applications in NEVs.