Paper presented at the 28th Annual Southern African Transport Conference 6 - 9 July 2009 "Sustainable Transport", CSIR International Convention Centre, Pretoria, South Africa.
An optimal smoothing algorithm is developed and applied to Moving SA (Johannesburg) data. Non-Linear regression is applied successfully to the Moving SA data. Values of 437.4 for saturation (vehicleslk.capita), 0.04569 for growth parameter (per annum frequency), and 1979.77 (years) for point of maximum growth are obtained. A correlation of +0.9647 with the raw data was achieved. The soundness of the statistical inference is demonstrated by showing the bias and trend of the model error term to be statistically indistinguishable from zero. Inner bounds on the confidence intervals for small sample non-linear regression are constructed for the predicted values.