Observer-based robust platoon formation control of electrically driven car-like mobile robots under collision avoidance and connectivity maintenance with a prescribed performance

Abstract

This study intends to address the platoon formation control problem of a team of N electrically driven underactuated autonomous car-like mobile robots. A platoon controller is proposed by using the relative distance and angle between each two successive robots in the platoon. A high-gain observer is also used to leave out velocity sensors to reduce the cost of the implementation/maintenance and the weight of the robots. Then, the dynamic surface control method is used to prevent complexity of the controller design. Next, by utilizing the prescribed performance bound method, predetermined desired transient and steady-state behavior of the tracking formation errors, robots connectivity preservation, and their collision avoidance are guaranteed as well as singularity avoidance. Adaptive neural networks and adaptive robust controllers are used to improve the tracking performance in the presence of parametric and nonparametric uncertainties, unmodeled dynamics, actuator saturation nonlinearity, and unwanted external disturbances including exogenous forces and torques, friction, and vibration. The Lyapunov direct method proves that all signals of the closed-loop control system are uniformly ultimately bounded. Finally, simulation results are demonstrated to show the superiority of the proposed convoy tracking system.

Integrated Geometric Optimal Control of Spacecraft Attitude and Orbit Based on SE(3)

“…There are many methods to constrain the control signal. For example, in [33]- [35] the saturation function is used to constrain the control signal, which is widely used and practical. In this paper, the constraint of the control signal is classified as one of the constraint equations of convex optimization based on MPCP convex optimization, and the calculation is more intuitive, concise and efficient.…”

Velocity‐free formation control for omnidirectional mobile robots with input saturation

“…Since | θ f (t )| < 𝜖 𝜃,e for any t ≥ 0, it follows that Equations (36) and (39) hold at all time. Hence, from the definition of Δ e in (36), we have that 1) If the disturbance 𝜹 f = 0 3×1 , then V e (t ) (33) and the associated estimation errors α f and β f will be exponentially stable.…”

“…In order to reduce the weight of mobile robots and the cost of sensors, an adaptive observer is proposed in [38] to estimate the unmeasured velocities, based upon the transformation matrices. In [39], a high‐gain velocity observer is also applied into the car‐like mobile robots, together with a low‐complexity dynamic‐surface‐based controller, the desire formation performance is achieved. Do and Pan [40] put forward a simple state transformation to facilitate the velocity observer design.…”

“…An observer with the extended Kalman filter is developed in [41] to estimate the velocities and disturbances of the mobile robots. The above results [16, 38–41] are derived for the nonholonomic mobile robots. For the OMRs, two generalized proportional integral (GPI) observers are proposed in [42, 43] to achieve velocity‐free tracking control.…”