Trajectory Tracking Control Of An Omnidirectional Mobile Robot

In the past few decades,omnidirectional mobile robots(OMRs)has attracted more and more attention in the field of wheeled robots because of its superior motion performance.Compared with the non-holonomic constrained robot,OMRs are complete constraint robots that can perform translational and rotational independently at the same time.This makes OMRs suitable for the work in narrow spaces,such as a hospital.This thesis has conducted an in-depth study of the trajectory tracking control of an omnidirectional mobile robot equipped with Swedish wheels.The Swedish wheeled OMRs are non-continuous nonlinear systems.The general control strategy for noncontinuous nonlinear systems is to make them continuous,linearized around the operating point,and convert them into a “continuous and linear” system for control.Firstly,this thesis establishes kinematics and dynamics model of a self-designed OMR prototype using Kiwi Drive platform and Swedish wheels.Also,a dynamic model using the average contact radius is deduced which makes the discontinuous system continuous.The unmodeled system dynamics,parameter uncertainties,and internal-external disturbances during continuation are estimated by the generalized proportional-integral(GPI)observer.A non-linear PD controller(NPD controller)is designed using feedback linearization method for the nonlinear system after GPI observer‘s compensation effect.The reliability and effectiveness of the proposed controller are verified,through simulation,experiment,and comparison with the control strategy containing the PD controller based on GPI observer.For the model-based control strategy,the system model needs to be established;there are unmodeled dynamic characteristics,the model parameters are not easy to be accurately obtained.The modified compact form dynamic linearization based modelfree adaptive control strategy is used to improve the trajectory-tracking performance of the OMRs based on the Kiwi Drive platform structure.Stability of the control system is analyzed.Its effectiveness is verified through simulations and experiments,and comparison with GPI-based PD controller and NPD controller.The controller has the advantages of model information free,simple structure,easy controller design,convenient parameter adjustment,a small amount of calculation and strong robustness.