Bee colony optimisation of the travelling salesman problem in light rail systems

摘要

The travelling salesman problem in transit planning is an extremely complex and non-deterministic polynomial problem. Many optimisation algorithms have been tested but failed to facilitate passengers getting to multiple stations by way of a complex rail transit network. Although various algorithms based on social insect species have successfully solved complex problems, bee behaviour in nature has inspired more significant solutions. The study reported in this paper aimed to identify the most efficient algorithm and develop a mathematical model based on artificial bee colony optimisation to solve this problem in light rail transit systems. For validation, the developed model was contrasted with empirical routes generated by human experience. The model was able to generate the optimal route in all cases without any error or miscalculation and could thus be used in any rail transit network in the world to determine the optimal route for multiple stations.