Statistical analysis of measurement data

The analysis of measurement data, including the consideration of measurement uncertainties, is a key component of metrological investigations. We offer statistical analyses of collected data sets for a wide range of applications. We use STATISTICA (StatSoft (Europe) GmbH) to carry out the data analysis.

A comprehensive range of methods is available, including:

• descriptive mathematical statistics (basic data preparation: calculation of statistical measures, including graphical representations, etc.)
• automatic classification (e.g. cluster analysis)
• supervised classification (e.g. discriminant analysis)
• time series analysis and forecasting

Project experience

As part of a research project, statistical measures were calculated for metrologically recorded consumption values (active electrical power, air and cooling lubricant consumption) of a machine tool in its various operating states. Furthermore, by applying various classification methods, machine states were examined for differences in their energy consumption behaviour. In the environmental sector, the impact of underground natural gas leaks on various road construction sites was determined, amongst other things, by using methods of automatic and supervised classification based on measurements of corresponding natural gas concentrations.

Sectors: Mechanical engineering, energy technology, environment and transport
References: Research projects in industrial measurement technology