A new reactive target-tracking control with obstacle avoidance in a dynamic environment

Abstract

This paper addresses a new reactive control design for point-mass vehicles with limited sensor range to track targets while avoiding static and moving obstacles in a dynamically evolving environment. Towards this end, a multiobjective control problem is formulated and control is synthesized by generating a potential field force for each objective and combining them through analysis and design. Different from standard potential field methods, the composite potential field described in this paper is time-varying and planned to account for moving obstacles and vehicle motion. Basic conditions and key properties are derived using rigorous Lyapunov analysis. Simulation examples are included to illustrate both the design process and performance of proposed control.

Secure practical tracking of an uncertain robotic system under coarse conditions on reference trajectory and obstacles

“…Specifically, a robotic manipulator is investigated in [19] without any uncertainties since unknown system parameters and disturbance are all excluded. In [20][21][22][23][24][25][26][27], several mobile robots are considered but external disturbance is excluded in [22][23][24][25][26][27] while the system parameters are completely or partially known in [20,21,24]. Moreover, some robotic systems described by general Euler-Lagrange equations or non-linear system are considered in [28][29][30], but the system matrices (or parameters) therein must be partially known while the disturbance has a known bound or must be matched with control input.…”

“…Besides the restriction on system uncertainties, the existing results on this topic are constrained by the generality of the reference trajectories as well as the measurability of the reference trajectories and obstacles. Specifically, the reference trajectories must be at least second-order continuously differentiable in [23,25,28,30] and hence exclude a large class of ones which are just first-order continuously differentiable or even non-smooth. In practice, the reference trajectories may be given by some moving objects whose trajectories are usually undesirable, and hence may be only first-order continuously differentiable or even nonsmooth.…”

“…Moreover, the time derivatives of both the reference trajectories and the obstacles in the first or second order (i.e. their velocity or/and acceleration) in [25,[28][29][30] must be available for feedback, but are difficult (or even impossible) to be obtained in practice. For example, in ocean engineering, one surface vessel tracks a moving target in certain environment with multiple dynamic obstacles which are given by other moving surface vessels.…”

“…(i) More serious system uncertainties are allowed. Different from [19][20][21][22][23][24][25][26][27][28][29][30] where system uncertainties are constrained (as aforementioned), this article allows all the non-linear dynamic matrices and mismatched disturbance to be unknown while without known bound, and hence allows more serious uncertainties. (ii) Coarse conditions are needed on reference trajectory and obstacles.…”

“…Different from [25,[28][29][30] where refined information are required on the reference trajectory and obstacles (as aforementioned), this article disregards their time derivatives and hence only requires coarse conditions on the reference trajectory and obstacles. (iii) A wider class of the reference trajectories can be followed under obstacle-avoidance.…”