Page 32 - Read Online

Sun et al. Intell Robot 2023;3(3):257-73 I http://dx.doi.org/10.20517/ir.2023.17 Page 259

firstdevelopedtomakeatrade-offbetweencommunicationefficiencyandcontrolperformance. The proposed
scheme, known as the state-sensitive (SS)-ETC scheme, incorporates the state measurement directly into the
adjustment of event-triggered threshold dynamically. The convergence of the dynamic event-triggered thresh-
olds is also removed while the stability of the path following control is not destroyed. The main contributions
of this paper can be summarized as follows:
• The SS-ETC scheme is developed to adaptively adjust the event threshold. A barrier-like function is first in-
troduced to indicate the SS-ETC scheme for path following control of AGVs. Compared to the static ETC
scheme [30] with a fixed event-triggered parameter, the proposed SS-ETC can adjust the event-triggered
threshold dynamically by introducing such a barrier-like function. Importantly, there is no need to intro-
duce extra terms or dynamics when adjusting the event-triggered threshold. This is very different from the
previous works, such as adaptive ETC scheme [25] and dynamic ETC scheme [26] .
• The state measurement will directly contribute to the adjustment of an event-triggered threshold. Unlike
the separate descriptions for physical dynamics and communication intervals in previous works [27,28] , the
dynamics of plant and communication scheme are intergraded in a uniform way. This approach gives
a clearer presentation on stability properties and communication behavior. In addition, the proposed SS-
ETC scheme is based on thesampled-data framework, which is different from thecontinuous one proposed
by [31] .
• The stability of the path following control of AGV is naturally guaranteed by the proposed SS-ETC scheme.
With the proposed SS-ETC scheme, all the event-triggered thresholds can be naturally suppressed within a
stability region. This approach provides a more flexible way to the co-design of communication and control
in path following control of AGV.
Compared to the previous works, the proposed SS-ETC scheme, based on state perception, shows a higher
level of adaptiveness on threshold adjustment in order to improve the functional safety of AGVs.

The remainder of this paper is organized as follows: Section II gives a more detailed description on the pro-
posed SS-ETC and NCS modeling for path following control of AGVs. Section III presents the main results
of this work, including the stability analysis and controller design under the proposed SS-ETC. In section IV,
simulation results, which were conducted on a networked path following control of AGVs, are shown to verify
the proposed theoretical results. Finally, section V concludes this paper.

To the end of this section, some physical meanings of related self-steering control of autonomous vehicles
(AVs) are given below Table 1.

2. PROBLEM FORMULATION
In this section, we first present the networked path following control modeling. Then, the SS-ETC scheme is
proposed to dynamically adjust the event-triggered parameter. At last, the path following control modeling of
AGVs under the proposed SS-ETC scheme is established.

2.1. Networked path following control framework
A networked path following control of AVs can be classified as a typical NCS, where all the measurements are
transmitted over a CAN bus and GPS network, among others. The control actions are calculated by an ECU.
To optimize the utilization of communication and computation resources, this paper introduces the SS-ETC
unit. The introduction of the SS-ETC unit aims to address the following two problems:

• Under the conditions of resource constraints of in-and-out vehicle networks, the SS-ETC unit should reg-
ulate its communication behavior to realize a more efficient utilization of resources.
• Considering the control requirements of the path following control, the SS-ETC unit must ensure the sta-
bility of the path following control of AVs under resource-constrained environments.