A cultural clonal selection algorithm based fast vehicle scheduling approach

Vehicle scheduling problem is very important for public transport management of bus companies. Given a bus timetable, the problem is to assign vehicles according to the timetable to minimize some optimization objectives. In this paper, a cultural clonal selection algorithm based approach is proposed to automatically obtain a vehicle scheduling solution. Firstly a set of candidate blocks is generated using initial start times. Then a cultural clonal selection algorithm is proposed to choose the best subset of blocks from the block set as a scheduling solution. An initial start time based antibody encoding scheme is suggested, which has the advantages of short coding and low complexity of decoding. An objective function is designed to maximize the occurrences of start times in the final solution. An adjusting strategy for departure times of vehicles is designed to improve the final solution. The proposed approach is applied to a real-world vehicle scheduling problem of the Bus Company of Xi´an city in China to evaluate its effectiveness. Experimental results show that the approach can quickly generate reasonable scheduling solutions, which fulfill the practical vehicle scheduling demands.