A Synchronous Model of Solving Train Scheduling Problem using Distributed Multi Agent System
Train scheduling is an essential problem of railways networks programming. The most common type of railways
is single-line which in turn creates more limitations. The efforts to solve this issue have been started since decades, the
previous and current efforts work by applying various techniques that contributed to solve the problem in different concepts.
However, despite the wide range of multi agent application and the complicated problems that solved by multi agent
systems, there were no significant implementation and use of multi agent systems to solve the train scheduling problem.
Therefore, we proposed a model for train scheduling in single-line railways, where we present a multi agent based model by
formulating this problem as a distributed constraint satisfaction problem (DCSP). Since in railways the trains travel the
different block simultaneously, hence the use of intelligent agents and provide a model that can solve the problem
simultaneously matches with the issue perfectly. Existing synchronization in our proposed method avoids of unnecessary
stop in trains travel and subsequently it gives an optimal scheduling. Experiments designed to evaluate the effectiveness of
the suggested solution; the implementation of the algorithm goes on two directions with and without constraints. The result
proves the multi agent systems ability to provide a valid contribution to solve the train scheduling issue.
Keywords - Intelligent agents; multi-agent systems; distributed constraint satisfaction problem; train scheduling.