Adaptive estimator‐based formation maneuvering control of nonholonomic mobile robots

Abstract

Summary In this paper, the formation maneuvering control problem for a group of nonholonomic mobile robots with the objective of having a desired formation shape described by distances between pairs of robots and an overall maneuvering velocity is studied. The desired maneuvering velocity, which can be constant or time‐varying, is only known to a set of agents. A control scheme consisting of an adaptive estimator and a modified gradient control law is proposed to solve this problem. The adaptive estimator is designed to estimate the desired maneuvering velocity in either constant or time‐varying situation. Utilizing the estimated velocity, a modified gradient control law is designed based on the nonholonomic kinematic model so that the objective is achieved. Local asymptotic convergence of the overall system is guaranteed by choosing appropriate control parameters. The effectiveness of the proposed control scheme is demonstrated through simulation results.

A novel saturated PID‐type observer‐based controller for wheeled mobile robots with a guaranteed performance considering path curvature

“…• The measurements of the velocity signals are not required for the proposed formation controller by designing a nonlinear velocity observer. In contrast to, [22][23][24][25][26][27][28][29][30] the proposed observer does not engage the robot dynamics while reducing unwanted peaks in the velocity signals estimation by using saturation functions.…”

“…Recently, many interesting theoretical results on the formation control of mobile robots have been reported by the researchers in the literature 12–21 . Based on a deep literature review by the author including, 1–33 the following shortcomings are evident in the previous works: According to a careful review of the literature including, 1–33 the majority of the literature is devoted to the formation control of differentially driven wheeled mobile robots which is a particular case of nonholonomic type ( m , s ) WMRs, that is, type (2, 0) robot. Since design of the previously proposed controllers is dependent on robot's kinematic model, they could not be applied to other types of mobile robots with different steerability, mobility, and kinematic models.…”

“…One way to remove such sensors from the robots is to make use of state observer schemes in the controller design process. After a deep literature review, there exist a few observer‐based controllers for the wheeled mobile robots which some of them has been reported in References 22–30. This shortcoming is due to the fact that the design of observer‐based controllers is a difficult and challenging problem in the presence of nonholonomic constraints, nonlinearities, and uncertainties. Many previous controllers including 1–33 assume that the robot actuators can accept every level of control input in practice and they often neglect the risk of actuator saturation which is unavoidable in practice.…”

Formation control of nonholonomic mobile robots with inaccurate global positions and velocities

“…In this article, perfect velocity tracking holds that followers' current velocities can track the desired velocities in practice when the time t → +∞, similar results can be found in References 33 and 37. Based on the auxiliary angular velocity (25) and the formation control law ( 26) and ( 27), the closed-loop error system is rewritten as…”

“…To improve the formation tracking accuracy, a precise kinematic model of the robot is important. In theoretical analyses [25][26][27][28] and experimental researches, [29][30][31][32] most kinematic models are established by using accurate global positions at robots' barycenters. However, the unknown robot's barycenter leads to the global positioning deviation, which generates unmodeled parts of the kinematic model.…”