image: The ETC-ACC control scheme addresses the practical issues of excessive communication traffic and low tracking accuracy by integrating composite learning, event-triggered technology, and asymptotic convergence control methods.
Credit: Yingjie Deng/Yanshan University,Fangcheng Liu/ Yanshan University,Songtao Wang/ Yanshan University, Yifei Xu/ Yanshan University, Tao Ni/ Yanshan University, Dingxuan Zhao/ Yanshan University
Researchers have developed an Event-triggered asymptotic composite neural tracking control scheme for intelligent vehicles, integrating radial basis function neural networks and a serial-parallel estimation model to compensate for system nonlinearities and uncertainties. Published in Advanced Equipment, this breakthrough has the potential to improve tracking precision of intelligent vehicles by ensuring asymptotic convergence of position error, enhance communication efficiency through event-triggered control, and boost system robustness against nonlinear dynamics and uncertainties, as validated by numerical simulations showing superior performance over traditional adaptive control methods.
High-precision tracking control for intelligent vehicles holds significant promise in advancing autonomous driving technology, yet it is challenged by system nonlinearities, uncertainties, and communication constraints that often lead to excessive traffic and low accuracy. Addressing this challenge, Professor Yingjie Deng from Yanshan University and Master's candidate Fangcheng Liu, in collaboration with researchers from the same institution, have developed an event-triggered asymptotic composite neural tracking control scheme that enhances both tracking precision and communication efficiency in intelligent vehicle systems. By integrating radial basis function neural networks, a serial-parallel estimation model, and integral-bounded functions, the scheme ensures asymptotic convergence of tracking errors while reducing communication traffic through variable threshold-based triggering conditions, with numerical simulations validating its superior performance in circular and abrupt trajectory scenarios.
“This control scheme marks a critical advancement in intelligent vehicle tracking," explains Professor Deng. "With reduction in communication traffic and a robust asymptotic convergence mechanism, this framework safely achieves high-precision tracking without compromising control efficiency, even in complex nonlinear driving scenarios."
"The newly developed control scheme delivers event-triggered updates, which are more communication-efficient than traditional continuous control methods. Its adaptive asymptotic convergence mechanism ensures compatibility with nonlinear vehicle dynamics, which are common in autonomous driving scenarios. Liu emphasizes, 'The event-triggered strategy significantly boosts communication efficiency and tracking accuracy, overcoming the challenges posed by system uncertainties and high-dimensional state spaces.'"
A major challenge in intelligent vehicle tracking is communication overload, which can lead to control delay and tracking deviation. The team addressed this by integrating a variable-threshold event-triggered mechanism and a serial-parallel estimation model into the control framework, reducing unnecessary communication updates during complex trajectory tracking.
While the team acknowledges the need for further research in addressing external disturbances and multi-vehicle coordination, this study represents a critical step toward more efficient and robust intelligent vehicle tracking control.
This paper ”Event-triggered adaptive neural network asymptotic tracking control of intelligent vehicles with composite learning” was published in Advanced Equipment.
Deng Y, Liu F, Wang S, Xu Y, Ni T, et al. Event-triggered adaptive neural network asymptotic tracking control of intelligent vehicles with composite learning. Adv. Equip. 2025(1):0002, https://doi.org/10.55092/ae20250002.
Method of Research
Computational simulation/modeling
Subject of Research
Not applicable
Article Title
Event-triggered adaptive neural network asymptotic tracking control of intelligent vehicles with composite learning
Article Publication Date
11-Jun-2025