| Initialization: initialize the corresponding parameters of discrete choice behavior model, and K. |
| Procedure: |
| if then |
| Define a temporary set List_k; |
| for do |
| if s==O then |
| Calculate the based on the Eq. (21). |
| else if s==D then |
| Calculate the based on the Eqs. (19) ~(20). |
| else |
| Calculate the based on the Eqs. (22)~(24). |
| if the number of elements in List_k is not exceeding K, add the into the set List_K and sort the elements |
| from smallest to largest; |
| if the number of elements in List_k equals K, and the last value of List_k is larger than , remove the last |
| value of List_k and add to the List_k. |
| end if. |
| end for. |
| Calculate the based on the using the Eqs. (25)~(26). |
| Then, calculate based on Eqs. (27). |
| Calculate the joint probability , , based on Eqs. (28)~ (30). |
| Using the Roulette method to generate an alternative travel decision: |
| (1) Construct a cumulative probability set R based on , , , where m=, |
| and . |
| |
| . |
| |
| (2) Generate a random variable r. |
| (3) Determine the behavior choices: |
| if , passenger will give up the metro journey and take a bus instead to the destination and ; |
| if , passenger will give up the metro journey and take the transport mode m instead to |
| the destination, ; |
| if , passenger will take an alternative station k as his new origin station of |
| metro journey, and he will take the transport mode m to the new origin station from the planned origin |
| station, the entry time at new origin is the original entry time added by the time consumption of the |
| transport mode m, ; |
| if , passenger will wait at the planned origin station. |
| else if then |
| Define a temporary set List_k, and we set the O as the passengers’ current station at the current time. If |
| passenger p is now in train, we set the O as the next station the train will stop at. |
| Judge whether the travel time from the current station to the destination is larger than the closure duration? |
| if yes, the passenger p need not to change his travel behavior. If not, go on executing the following part. |
| for do |
| if s==O then |
| Calculate the based on the Eqs. (19)~ (20). |
| else if s==D then |
| Set , where W is a very big constant. |
| else |
| Calculate the based on the Eqs. (22)~(24). |
| if the number of elements in List_k is not exceeding K, add the into the set List_K and sort the elements |
| from smallest to largest; |
| if the number of elements in List_k equals K, and the last value of List_k is larger than , remove the last value |
| of List_k and add to the List_k. |
| end if. |
| end for. |
| Calculate the based on the using the Eqs. (25)~(26). |
| Then, calculate based on Eq. (27). |
| Calculate the joint probability , , based on Eqs. (28)~(30). |
| Using the Roulette method to generate an alternative travel decision: |
| (1) Construct a cumulative probability set R based on , , , where m=, |
| and . |
| , m= |
| , m=, . |
| (2) Generate a random variable r. |
| (3) Determine the behavior choices: |
| if , passenger will give up the metro journey and take a bus instead to the destination from current |
| station, ; |
| if , passenger will give up the metro journey and take the transport mode m instead to |
| the destination from current station, ; |
| if , passenger will take an alternative station k as his new |
| destination station of metro journey, and he will take the transport mode m to the planned destination |
| station from the new destination station, ; |
| end if. |
