Text Box: Error and uncertainty Short course   Introduction  Error and uncertainty inevitably arise in the process of modelling real-world problems. This course identifies the common sources of error and uncertainty that arise in the modelling, simulation and analysis processes, and discusses practical means of error and uncertainty estimation and treatment. The course is aimed at both designers and engineers who wish to understand where, how, and why errors and uncertainties arise, and how they can be dealt with. The course begins by discussing the potential sources of error and uncertainty that can arise in the simulation modelling process, and explains how they can be classified and treated. Idealisation error arises during abstraction from the real-world and is manifested in the form of invalid modelling assumptions which introduce error into the equations and data. Numerical error principally arises during conversion of the conceptual model from the continuous to the discrete domain, and can take the form of discretisation, truncation, and round-off approximation. Numerical error also arises during solution processing of the model equations. Uncertainty arises where it is not possible to define parametric values accurately, such as in the modelling of natural loads and materials. Finally, techniques for model verification and validation are explained, together with methods for correlation with common sources of corroborative data. Course content  Introduction What are the common sources of error and uncertainty in modelling? How can concepts of error estimation and treatment be applied? Sources of modelling error Idealisation error Numerical error Idealisation error Model equations Model data Numerical error Discretisation error Truncation error Round-off error Error estimation Spatial discretisation error Temporal discretisation error Error analysis and convergence Practical error estimation and adaptivity Uncertainty estimation Sensitivity analysis Uncertainty bounding Model correlation Experimental Experiential Alternative modelling Conclusions Model improvement

Presenter:      Jonathan Smith

Duration:          Half day, p.m.

 

Synopsis

Error and uncertainty inevitably arise in the process of modelling real-world problems.

This course identifies the common sources of idealisation and numerical error that typically arise in the modelling, simulation and analysis process, and discusses practical means of error identification, estimation and treatment.

The course is aimed at both designers and engineers who wish to understand where, how and why errors arise, and how they can be dealt with.

© Copyright 2009 Compusis

Short courses

 

Finite element method

 

Error and uncertainty

 

Modelling, verification and validation

 

Quality management

 

 

 

Course schedule

 

Course registration

 

 

Testimonials

 

Presenter

 

 

 

computational modelling, simulation and analysis

compusis

Consulting

Contact

Short courses

Home