Energy-Efficient Local Path Planning of a Self-Guided Vehicle by Considering the Load Position

The local path planning, as one of the navigation stages, plays a significant role in the energy consumption of Self-Guided Vehicles (SGV). Since SGV must operate for several hours on a single battery charge to transport loads, its energy consumption is a critical issue. Therefore, this article puts forward an approach for boosting the energy efficiency of the local path planning stage using load position. Unlike other similar works which solely use robots' kinematic and kinetic constraints to develop energy-efficient local path planners, this article considers the effect of load position on SGV's dynamic. In this regard, first, the kinetic model of the differential drive SGV is developed to consider the change of SGV's Center of Mass (CoM) affected by load properties. Second, machine learning methods are used to create two learning models for online estimation of the position of CoM (PoCoM) and prediction of required energy of sample trajectories. Hence, the generated SGV's kinetic model is used to train the learning models. Finally, estimated parameters are employed to add a new constraint to extend the cost function of the local path planner. The outcomes of the study show that the proposed planner generates smoother and shorter paths to pass obstacles and corridors than a general one. Thus, SGV's energy consumption decreases by considering the load effect.