Search Results

Careful experimental design ensured that the EPEFE programme would deliver accurate, unambiguous conclusions despite the many pervading constraints on test fuels and test order. Fine-tuned statistical and graphical procedures were used methodically to identify possible outliers in the emissions data collected and to correct for underlying systematic trends. Summarising and interpreting the validated data required the rigorous use of advanced statistical modelling techniques which accounted correctly for both short- and long-term variability in emissions, and any unavoidable non-orthogonality or imbalance in the design. Multivariate methods were used for further data exploration.