There is a missing attribute in the chromosome in genetic algorithm

I'm building a model about airline scheduling by using genetic algorithms that generate a flight schedule on a daily basis and assigning the right aircraft too. Each chromosome represents a total of 50 flight legs a day(including the return trip that means 100 trips a day) and I have 42 legs, so obviously there will be repeated flight legs. The initial members are created randomly, and there are some constraints that must be satisfied to have a valid schedule. The problem is one of the constraints is that all flight legs must be included, but the model always fails to achieve that!! It starts by missing like 9 or 10 destinations, then it decreases by using creep mutation until it reaches 1 and gets stuck in it and never finds it! Even when it runs for 1000 generations! I tried to interference in the initial member and include them all before the beginning but that doesn't work either! so what's the matter?! how could I solve this?