Fault Detection in Cables: A Machine-Learning Approach

Fault location identification is one of the most common, but challenging, activities that power operators face within industrial plants. Once a fault occurs in the network, protective relays isolate the affected area from the system; then, maintenance crew use special tools and patrol the area to pinpoint the fault. This process may take a lot of time and effort depending on the fault type, detective tools, and cable position. In order to address this issue, the current project seeks to develop an advanced hardware-software package, which will be developed based on machine learning and to be capable of finding exact fault location in a fraction of second. The reported output of the device can be either used as the backup hints, or as the primary detection tool in the foreseeable future to help the operators in different mining plants.

Faculty Supervisor:

Chi Yung Chung;Seyed Mahdi Mazhari;Ha Nguyen;Seok-bum Ko

Student:

MD Salauddin

Partner:

Cameco Corporation

Discipline:

Other

Sector:

University:

University of Saskatchewan

Program: