Research Article

First-Train Timetable Synchronization in Metro Networks under Origin-Destination Demand Conditions

Algorithm 1

NSGA-II.
Step 1: (Initialization)
  Step 1.1: Input the number and index of the lines in the network and the number and index of the intermediate stations and transfer stations on each line;
  Step 1.2: Input OD demand set K and promising routes;
  Step 1.3: Input the actual schedule of each line in the network and other parameters;
  Step 1.4: Input algorithm parameters, such as the population size pop_size, maximum number of iterations Gen_max, crossover probability, and mutation probability;
Step 2: (Initial population)
  For i = 1 to pop_size do
    Generate l_num departure times of first trains and headways of each individual randomly;
  Endfor//i
  Generate the parent population;
  Gen = 1;
Step 3: (Sorting)
  Step 3.1:
    For j = 1 to pop_size do
      Calculate the nondominated sorting rank of each individual;
    Endfor//j
  Step 3.2:
    Calculate the crowded distance of each individual in the same rank;
Step 4: (Genetic operation)
  According to the binary tournament selection rules, pop_size/2 individuals are selected from the parent population and the pop_size/2 individuals are genetically operated according to the randomly generated crossover probability and mutation probability until other pop_size/2 individuals are generated;
Step 5: (Offspring population)
  These pop_size individuals generated in Step 4 compose the offspring population;
Step 6: (Combined population)
  The parent population generated in Step 7 (if Gen = 1, generated in Step 2) and offspring population generated in Step 5 compose the combined population (2pop_size);
Step 7: (New population)
  Calculate the nondominated sorting rank of each individual in the combined population and calculate the crowded distance of each individual in the same rank;
  Select pop_size new individuals from the combined population to compose the new parent population;
Step 8: (Solution)
  If Gen<Gen_max, then
   Go to step 4;
Else
   Output final solutions, such as the Pareto front, synchronized timetable, and train connection sequence.
Endif