Trip-production rates presented in cross-classification tables are essential data for the planner's understanding of the travel characteristics of a region. Trip rates obtained from surveys, however, often show a pattern that is not consistent with what is expected by the analyst; for example, the greater the household size and auto ownership, the greater the number of trips generated. This pattern may not be found in the trip rates that are obtained directly by the survey. In such cases, analysts commonly adjust the irregularities manually. The way in which the values are adjusted affects the credibility of the trip table and, ultimately, the forecast travel demand. A method that adjusts the values of the trip table systematically is presented. The process uses the fuzzy linear programming method. The objective is to make the adjusted value as close to the observed value as possible. The constraints are to make the adjusted values adhere to the analyst's general expectations about the pattern of the values in the table, and to match the number of trips estimated from the adjusted trip table with the actual number of trips surveyed. An application example that uses real-world data is given.
Abstract