education

Education


Round Corner
Department of Computer and Information Science

Oppgaveforslag

Application of machine learning to ILI data denoising

The project is for Master thesis work at AkerSolutions

Objective:

The object of this study is to apply machine learning technique on ILI data denoising and perform case studies to the remaining capacities of corroded pipeline. The basis of this study shall be in line with the DNVGL-RP-F101 (2017).

 

Background:

Aker Solutions AS has developed in-house software rbpGen and corrPipe to process the survey data (ILI, in-line inspection data) and estimate the remaining capacities in 2017. The ILI data could be provided by Statoil ASA for one of its operating field. The screening of the survey data is challenging due to large amount of ILI data. And consequently, it will have impact to the remaining capacities check. The software developed is in close agreement with DNVGL-RP-F101. However, it may be not able to give the “correct” results for those highly scattered results. In this respect, advanced methods other than the recommendations in DNVGL-RP-F101 might be needed, such as the machine learning, image denoising et al.

 

Scope of work:
The scope of work may consist of:

  • Be familiar with the latest DNVGL-RP-F101 (2017);
  • Be familiar with the rbpGen and corrPipe software;
  • Be familiar with the Python language, which is the main scripting language used in Aker Solutions SURF, Trondheim;
  • Review of the data filtering methods available in machine learning; The investigation of the possibilities of using machinery learning technique on processing the ILI data;
  • Case study based on the ILI data provided by Statoil ASA.
  • Concluding the case study and further recommendations to the software development.

 

Skills required:
Python, statistics theory, machinery learning, pipeline engineering, solid mechanics et al., DNVGL-RP-F101 (2017).

 

Reporting:
A time schedule should be presented in written from at project meeting if necessary. The work should be presented in written form, common rules of reporting technical work should be met.

 

Supervisor at NTNU: Prof. Zhirong Yang

Co-supervisor at AkerSolutions: Dr. Zhenhui Liu

Faglærer

Zhirong Yang Zhirong Yang
Professor
320 IT-bygget
735 93440 
 
NTNU logo