Bike sharing systems without docking stations have got much popularity and rapidly growing in today’s era and becoming an integral part of green Transportation systems but in parallel,it raises severe management and rebalancing issues,which are becoming a challenge in their development and sustainability.Bicycles availability and shortages are in direct relation with the user satisfaction level.A uniform distribution of bicycle inventories over a network of zones to achieve optimal inventories with minimum cost is mainly focussed in this study.Rebalancing problem of dockless sharing bicycle’s systems is addressed in two stages.In the first stage,the randomly scattered bicycles positions are traced out in an area where rebalancing activity will take place and then further divided them into several groups or clusters of either surplus or deficit ones through the clustering technique of K-means clustering algorithm which is a distance-based algorithmic approach that minimizes the distance of bicycles and assigned bicycle to cluster.These clustering and collection mechanism will develop the supply and demand level in each zone which could be thoroughly manipulated before starting the main truck rebalancing operation among zones.
　　In the2nd stage of Intra zonal rebalancing,a heavy vehicle or Truck routing model is incorporated using Mixed Integer Linear Programming technique with a set of two homogeneous vehicles from two depots to all other collection points of each zone in order to pick up and to further deliver the bicycles at required deficit areas for attaining an equilibrium of inventories over the zones up to an optimal level.This will enhance the system towards an equivalent distribution of bicycles over the zonal network resulting into a greater users satisfaction level in terms of bicycles availability.Aftermath the developed Mixed Integer Linear Programming model is further tested using IBM-ILOG CPLEX Solver,for a different set of data to analyse its impacts by varying the supply and demand of bicycles of the zones on the user dissatisfaction level and checked the model’s validity and rationality.