A prescribed performance PID control of robotic cars with only posture measurements considering path curvature

This paper addresses the prescribed performance adaptive neural observer-based PID controller for robotic cars by considering the trajectory curvature compensation. Constrained relative distance and angle errors between a car and leader are transformed into a new Euler-Lagrange model of unconstrained errors via prescribed performance technique. Two filtered PID error variables are introduced to propose a neural adaptive robust observer-based controller which obtains the followings: (i) the prescribed performances of tracking errors are ensured, (ii) the controller singularity is avoided, (iii) the trajectory curvature is compensated to avoid unwanted deviations between the leader and follower, (iv) car velocities measurements are not required, and (v) controller compensates uncertainties by combining neural networks and adaptive robust control. Lyapunov's method proves the controller stability. Comparative simulations verify controller advantages. (c) 2022 European Control Association. Published by Elsevier Ltd. All rights reserved.

Automated summary

This study proposes a PID controller based on prescribed performance adaptive neural observer for robotic cars, with the consideration of trajectory curvature compensation, which ensures stability and advantages of the controller.