The Development of Linear Distance Functions for Distribution Analysis

The general method of estimating the road mileage between points identified only by their co-ordinates is to proportionately inflate the straight-line distance between them. We present the traditional single-region model and extend this to a generalized multi-region format. The inflating parameters are derived and various statistical tests carried out to evaluate the models. It is shown that, on average, the multi-region approach gives better mileage estimates for all types of distribution analysis. For specific problems, which are discussed, the single-region model is shown to produce results which are unacceptable for the planning of distribution networks.