Dual Redundant Flight Control System Design for Microminiature UAV

Abstract

Evaluation of varios soil erosion models with large data sets haveconsistently shown that these models tend to over-predict soil erosionfor smalm measured values, and uder-predict soil erosion for largermeasured values. This trend appears to be consistent regardless ofwhether the erosion value of interest for individual storm, annualtotals, or average annual soil losses, and regardless of whetherthe model is empirical or physically based. The hypothesis presentedherein is that this phenomenon is not necessarily associated withbias in model predictions as a function of treatment, but ratherwith limitations in representing the ramdom component of the measureddata within treatments (i.e., between replicates) with a deterministicmodel. A simple exemple is presented, showing how even a "perfect"deterministic soil erosion model exhibits bias relative to smalland large measured erosion rates. The concept is further tested andverified on a set of 3007 measured soil erosion data pairs from stormsof natural rainfall and run-off plots using the best possible, unbiased,real-world model, i.e., the physical model presented by replicatedplots. The results of this study indicate that the ommonly observedbias, in erosion prediction models relative to over-prediction ofsmall and under-prediction of large measured erosion rates on inividualdata points, is normal and expected if the model is accurately predictingerosion rates as a function of environmental conditions, i.e., treatments.