Robust Optimization of Travel Management in Europe
Charles Chen
Applied Materials
Mason Chen
Stanford University
This paper was addressing travel cost & time while visiting five major cities in Europe. The main objective was to build a robust route which can meet both time & expense requirements. The input variables used were the intra-city time & expense while the response variables were the total time & expense. Several travel design constraints were considered in the model as noise factors. Each input variable was limited to 2 different transportation choices: flight or train. The literature research was conducted about train & flight speed in Europe. Taking train is better when travel distance is under 500km, and flight is better if above 1,000 km. The choice can go either way in between. Most major cities in Europe are in that range. A multiple regression model was built on the mean time & expense. Optimal travel route was set by meeting the desirability functions of 2 travel requirements. To achieve Robust Design, Monte Carlo (MC) Simulation was conducted by including the time & expense variation. MC results have observed 8% risk of not meeting the expense budget. By raising the importance of time desirability, the new optimal route can meet both time & expense requirements.