Car-sharing is a new concept which enables users to rent cars for short period of times. Car-sharing systems can be classified as one-way or round-trip according to restrictions imposed for pick-up and drop-off locations. In more restricted round-trip systems, users have to return vehicles where they have picked them up, whereas one-way systems allow users to return vehicles to different drop-off stations. In addition, one-way systems have two types. If the car-sharing system operates with designated parking locations it is characterized as non-floating. If the users are allowed to drop-off vehicles with defined borders then the system is called free-floating.

In this research we are dealing with operational decisions of one-way non-floating electric car sharing systems with reservations and dynamic relocations, i.e. relocation can be done anytime during the operations. Our previous work in this area is addressing issues related to strategic and operational problem , and simulation of relocation operations. In this paper we are introducing the following new modelling concepts: (i) station clustering and (ii) integration of optimization with simulation for the operational problem.