Design sensitivity in the presence of uncertainties using Entropy and Fisher information Seminar

3D computational models have been widely used to support structural design, but data plays an increasing role in the presence of uncertainties. In this talk, I will present a new sensitivity metric using the Fisher Information Matrix (FIM) so that high value of information from uncertain data can be identified to support design decisions.

The talk will consists of mainly three parts.

• First, I will show that not only the FIM is closely linked to Entropy that is commonly used to quantify design complexity, the trace of the FIM also provides an upper bound of the sensitivity of failure probabilities for reliability engineering.
• Second, the role of the proposed metrics in engineering design process is discussed.
• I will then finish the talk off with a case study on the dynamic design of offshore structures subject to random wave excitations.

The work presented here is funded by the EPSRC grant ‘ Digital Twins for improved dynamic design ’.

Speaker : Dr Jiannan Yang, University of Cambridge

Jiannan Yang is a Postdoc at the Dynamics group, Engineering department, University of Cambridge. He received his PhD from ISVR under the supervision of Prof David Thompson and the research topic was on simulation of railway impact noise. After his PhD, he has worked for Hitachi in both Japan and the UK. At Hitachi, Jiannan worked on various rolling stock projects for noise and vibration control, including the Azuma trains that went into passenger service on the east coast main line in 2017. Since 2018, he has joined the EPSRC funded project on digital twins at Cambridge and he is looking at the use of computational vibration models in the engineering design process.

Microsoft Teams meeting

Join on your computer or mobile app

Click here to join the meeting

Or call in (audio only)

+44 20 3794 0272,,627279451# United Kingdom, London

Phone Conference ID: 627 279 451#