INFORMATION
- Research Name Novelty Detection and Autonomous Monitoring of Locomotives
- Client Peaker Services, Inc.
- Category PRACTICUM PROJECTS
- RESEARCH YEAR 2019
Project overview
This project focuses on developing novelty/anomaly detection methodologies based on locomotive engine operating conditions and associated statistical trends and relationships of over 600 sensors’ data collected which could also provide a basis for autonomous monitoring of the locomotive.
The methodology was implemented using two approaches -The autoencoder (artificial neural network) and Multivariate Statistical Process Control. The aim of an autoencoder was to learn a representation (encoding) for a set of data, typically for dimensionality reduction, by training the network to ignore signal “noise”.
The second method uses “Multivariate Statistical Process Control” which refers to a set of advanced techniques for the monitoring and control of the operating performance of batch and continuous processes.
Student team
Anupam Banerjee, Devon Ankar, Joel Woznicki, Krishna Priya Bandi, and Laxmi Siva Tejaswi Baipa
