Experimental validation of an autonomous control system on a mobile robot platform

Abstract

An autonomous control system designed for a non-holonomic wheeled mobile robot that is programmed to emulate a fixed-wing unmanned air vehicle (UAV) flying at constant altitude is experimentally validated. The overall system is capable of waypoint navigation, threat avoidance, real-time trajectory generation and trajectory tracking. Both the wheeled mobile robot experimental platform and the hierarchical autonomous control software architecture are introduced. Programmed to emulate a fixed-wing UAV flying at constant altitude, a non-holonomic mobile robot is assigned to follow a desired time-parameterised trajectory generated by a real-time trajectory generator to transition through a sequence of targets in the presence of static and popup threats. Hardware results of the autonomous control system where the trajectory tracker applies two velocity controllers accounting for fixed-wing UAV-like input constraints, are compared to simulation results of dynamic controllers that are based on non-smooth backstepping to demonstrate the effectiveness of the overall system.

Trajectory tracking of wheeled mobile robot by adopting iterative learning control with predictive, current, and past learning items

“…According to inequalities (21), (22), and (23), we can make the conclusions that when the parameters b β = 0, b γ = 0 and b q 0 = 0, the variables ũ i α , q i α , and ỹ i α all converge to zero. Otherwise, the variables ũ i α , q i α , and ỹ i α converge to bounded values which are the functions of the parameters b β , b γ , and b q 0 .…”

“…20 Also, it is the key technology to realize the fully autonomous motion and intelligence of robots. 21,22 Since the wheeled mobile robot is featured with time-varying, nonlinear, and strong-coupling dynamics properties, with the influence of measurements inaccuracy, as well as external disturbances and load changes, so actually it is rarely possible to obtain the complete and accurate model of a wheeled mobile robot. 23 The iterative learning control algorithm will determine the expected linear and angular velocities for the wheeled mobile robot, while the dynamic controller will control the actual speed and make it try to track the desired speed.…”

Cascade design for formation control of nonholonomic systems in chained form

“…However, a lot of interesting applications involve dynamical systems that are purely nonlinear and which exhibit some geometric constraints or henceforth called nonholonomic constraints for mobile robots [34]. The cooperative control of nonholonomic robot are considered in Ren [35], Ren et al [5] where the authors transformed the dynamic of the robots to double-integrator dynamics which gives the ability to control the hand position instead of the inertial position of the robots and therefore, the robots' heading cannot be controlled. It seems then desirable to have control over all the state vectors of each mobile robot in the formation.…”

“…Suppose thatx à e ¼ 0 then the third equation in Eq. (47) disappears and the system (47) is nothing but the cascade (35) analyzed in the proof of Theorem 4.3. This allows to conclude about objectives (16)- (18).…”

Globally convergent leaderless formation control for unicycle‐type mobile robots

“…One of the research topics is formation control, which implies the robots are driven to fulfil a specified geometric pattern. A considerable number of literatures have investigated the formation problem for unicycle‐type mobile robots [1–7]. Liu and Jiang in [8] study the leader–follower formation problem, in which the model of the robot is transformed into a double integrator by feedback linearisation.…”