In this report, the characteristics of the applied on-site measurement methods were explored based on theoretical study. According to the comparison of the five methods, it is indicated that the radar and GPS are easier to measure the speed; however, compared with the radio and mirror measurements, the external environmental factors could be hard to be avoided so that the unexpected speeds exist in the data processing. Moreover, the attributes and distributions of each measurement have been achieved applying statistics cooperated with relevant software. Comparing the results of histograms and goodness-of-fit tests of the methods, it could be concluded that the speed data of ATC are more representative of the traffic condition due to the better normalcy. In the conclusion, several suggestions are presented: (1) the unexpected speeds were observed because of manual or mechanical problems, therefore the operations could be strictly in accord with the requirements of each measurement; (2) during the data processing, the statistics should be applied properly, especially the determination of sample size should be taken more consideration, and the coefficients should be selected in terms of different requirements of the analysis. (Author/publisher)
Abstract