Adaptive coordinated collision avoidance control of autonomous ground vehicles

Abstract

This article presents a novel coordinated nonlinear adaptive backstepping collision avoidance control strategy for autonomous ground vehicles with uncertain and unmodeled terms. A nonlinear vehicle collision avoidance vehicle model which describes the coupled lateral and longitudinal dynamic features of autonomous ground vehicles is constructed. Then, a modified artificial potential field approach which can ensure that the total potential field of the target is goal minimum, is proposed to produce a collision-free trajectory for autonomous ground vehicles in real-time. Furthermore, in order to handle with the features of coupled and parameter uncertainties of autonomous ground vehicles, an adaptive neural network-based backstepping trajectory tracking control approach is proposed for collision avoidance control system of autonomous ground vehicles, and the stability of this proposed control system is proven by the Lyapunov theory. Finally, the co-simulation and experimental tests are implemented and the results show that the proposed collision avoidance control strategy has excellent tracking performance.

Control Strategies on Path Tracking for Autonomous Vehicle: State of the Art and Future Challenges

“…To solve the problem of chattering on path tracking based on SMC, a vehicle following control strategy based on adaptive SMC control is proposed considering load transfer characteristic [105][106][107][108][109][110][111][112][113][114][115][116][117]. Chuan Hu et al proposed adaptive SMC control strategy based on combination composite nonlinear feedback control (state feedback and state deviation feedback) and radial basis function neural network considering the effect of vertical motion on lateral velocity to improve tracking transient response and robustness [118].…”

Fault detection and isolation based on space projection operator for hypersonic re-entry vehicles with concurrent actuator faults

“…The admissible deflection angles of the aerodynamic surfaces are given in Table 2. 8 When actuator faults occur in the HRV system, the control allocation problem is addressed by minimizing the following indicator…”

A motivational driver model for the design of a rear-end crash avoidance system

“…Engineering driver models or controllers try to present computational input–output models to guarantee stable vehicle control for specific applications. There are some examples of driving models in control engineering, such as car following models, 7 collision avoidance model, 8–10 predictive control models, 11 fuzzy logic models, 12 sliding mode control, 13 neural network models, 14 and optimal control models. 15,16 These models and controllers are not psychologically plausible; thus, they cannot be used effectively for human–machine interactions.…”

“…Combined lateral and longitudinal collision avoidance systems need motion planning methods to design a path to avoid collision with moving or fixed obstacles. Some researchers use potential field, 8,10 model predictive control, 9 optimization, 15,16 and polynomial curve fitting 17 for combined lateral and longitudinal collision avoidance systems. Motion planners still need to improve real-time computational complexity, path continuity, vehicle kinematics and dynamics consistency, exactness of solutions in tight spaces, and consideration of human driver interaction.…”