A naukri.com initiative
Arxiv
2d
155
Image Credit: Arxiv
Machine Learning-Based Classification of Oils Using Dielectric Properties and Microwave Resonant Sensing
- The paper presents a machine learning-based approach for classifying oils using dielectric properties and a microwave resonant sensor.
- Oils exhibit unique dielectric behavior influenced by their molecular composition, resulting in specific changes in the sensor's resonant frequency and amplitude response.
- Variations in sensor responses are analyzed to extract relevant features used as inputs for various machine learning classifiers.
- The microwave resonant sensor functions in a non-destructive, low-power mode, ideal for real-time industrial applications.
- A dataset is created by altering oil samples' permittivity and recording sensor responses for training and evaluation.
- Multiple classifiers are developed and tested using the extracted resonant features to differentiate between types of oils.
- Experimental outcomes reveal a classification accuracy of 99.41% with the random forest classifier, indicating the method's efficacy in automated oil identification.
- The system's small size, energy efficiency, and high accuracy emphasize its suitability for rapid and dependable oil characterization in industrial settings.
Read Full Article
9 Likes
For uninterrupted reading, download the app