Autonomous Navigation Control Based On Improved TEB Algorithm

As one of the key technologies of mobile robots,autonomous navigation has attracted a lot of attention.In particular,local path planning and navigation control are the two most prominent problems.However,local path planning algorithms still have poor real-time performance especially implemented in low-cost micro-computers,and navigation control algorithms are sensitive to the robot's moving speed as well as sensor noises,resulting in low control accuracy.Therefore,this thesis focuses on these issues and studies autonomous navigation control of mobile robots.For the dynamic path planning of mobile robots,this paper first improves the TEB(Timed Elastic Band)algorithm,and proposes a method of reducing the number of optimization variables and using parametric cubic spline interpolation to interpolate path points into a smooth path curve.The improved method effectively reduces computing cost of the TEB algorithm in a dynamic environment,and addresses the issue regarding real-time performance.At the same time,this thesis proposes a method of using ESO(Extended State Observer)based discrete state observer to effectively estimate the running state of dynamic obstacles,which together with the improved TEB algorithm solves the dynamic path planning problem in presence of moving objects in the working space.Considering the problem of navigation control of mobile robots,this thesis firstly proposes a TEB algorithm based on model predictive control.The first-order inertial link is used to simulate the dynamic characteristics of the driving motors of a mobile robot,which effectively improves the transient response of navigation control compared to the original TEB algorithm.However,due to its poor performance when it is close to the target point,this thesis then proposes an improved line-of-sight-based pure tracking algorithm,which introduces deviation control of heading angle,and the algorithm has nothing to do with motion model of mobile robots,which effectively solves the vibrating problem and improves robustness in the case of uneven ground.