Engineering statistics: Normal, exponential and Weibull distribution, confidence interval.
Statistical experimental design: physical experiments and simulations, censored and suspended test.
Probabilistic design; Monte Carlo simulations (with Matlab and Ansys) of variation of performance caused by variations in design - manufacturing tolerances, material properties, geometric configuration), user (anthropometric data) and environment parameters (humidity, electromagnetic fields, temperature, dirt).
Robust design; minimise performance variations that is caused by variation of design sparameters, human properties and environment conditions.
A student who has completed the course should be able to:
• describe characteristic product properties in statistical terms
• estimate the confidence interval for the estimated reliability of a system
• define type of probability distribution for a given amount of data
• describe aim, methodology and result of a statistical experimental design
• create a test plan for a physical and a numerical experiment,
• describe aim and procedure to carry out a Monte Carlo simulation,
• use Monte Carlo simulation to analyse how uncertainty in the model parameters influences the simulation result
• describe the aim for robust design and how the method relates to optimisation methods
• use robust design to decrease the sensitivity of the performance of a product for variations in the parameters of its components
• use robust design to decrease the sensitivity of the performance of a product for variations in the technically interactive parameters of the system
• use robust design to decrease the sensitivity of the interactive performance of a product for variations in the ergonomic parameters of the system.