摘要：AbstractThe many advantageous features of GPS-based longitudinal surveys associated with prompt recall surveys make such surveys very attractive for travel behaviour studies. However, the sample size calculation procedure for GPS-based surveys is more complicated compared to well-known and widely applied conventional household one/two-day travel surveys. The higher cost of GPS surveys requires scrutiny at the sample size planning stage to ensure cost effectiveness. The essence of sample size calculations problem is of a trade-off between cost/time taking the precision of the survey into account. Different machine learning-based techniques have been developed to infer the transportation mode based upon speed and acceleration calculated from GPS data. However, none of these studies calculate the sample size required for validating these techniques. Calculating the most effective sample size for this inference mainly depends on the variability of these variables which are normally used. To perform this calculation, we develop an understanding of inter-modal (variation between different transportation modes) and intra-modal variability (variation within each transportation mode). The study demonstrates that the motorised modes reflect the highest variability. We use traffic count data to study this variability across different seasonal divisions. The hourly and daily seasonal divisions are proved to be of the highest variability. Extending the survey length also decreases the sample size significantly. This reduction is applied to the calculated sample sizes defining the survey length to be 2 weeks, taking the weekly-seasonality into account.
