Algorithm 1. Designing Shuttle Bus Routes with Genetic Algorithm

Input: generated shuttle stops, distance matrix, iteration epochs maxiter, crossover probability pc, mutation probability pm, population size pop_size, population selection probability p_select. 1Initialize populations to the default, i.e., generate random sequences for all shuttle stops. 2Based on Constraints 2-1 to 2-3, split and generate the initial sets of shuttle routes. 3Calculate the length and fitness of the initial route set, fitness = 1/sum of routes length. 4Randomly select an initial set of shuttle routes as the best chromosome. 5 Loop for maxiter iterations, in each iteration: 5.1 Population selection, retaining populations (shuttle routes sets) with high adaptation, i.e., route sets with short mileage. 5.2 If crossover conditions are satisfied (pc and pop_size), do crossover between chromosomes of two populations: 5.2.1 Randomly select two crossover stops in the range “2” to “No of stops”. //Note: The first stop is fixed, namely Shengxianhu Station. 5.2.2 Perform chromosome crossover and ensure that there are no duplicate stops in the same set of shuttle routes. 5.3 If mutation conditions are satisfied (pm and pop_size), do mutation for each chromosome. 5.3.1 Choose two random numbers [a], [b] in the range “2” to “No of stops”. 5.3.2 Swap the values of new gene [a] and new gene [b]. 5.4 Based on Constraints 2-1 to 2-3, split and generate a new set of shuttle routes. 5.5 If the fitness value of the new route set is greater than that of the best chromosome, then set best chromosome = new chromosome. 6End of Loop. Output: The best set of shuttle bus routes.