Event-triggered Iterative Learning Control for Perfect Consensus Tracking of Non-identical Fractional Order Multi-agent
- PDF / 2,168,330 Bytes
- 17 Pages / 594.77 x 793.026 pts Page_size
- 84 Downloads / 187 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
Event-triggered Iterative Learning Control for Perfect Consensus Tracking of Non-identical Fractional Order Multi-agent Systems Liming Wang and Guoshan Zhang* Abstract: This paper is devoted to the perfect tracking problem of output consensus for a class of non-identical fractional order multi-agent systems (NIFOMASs), in which different agents have different and unknown factional orders and dynamic functions. For the NIFOMASs including one leader agent and multiple follower agents, by designing the event-triggered mechanism along an iteration axis and introducing it into the iterative learning controller, an event-triggered iterative learning consensus protocol is proposed to reduce the number of controller update and to save the communication resource. By analyzing the convergence of learning process, the sufficient conditions are derived to guarantee that the output consensus tracking can be perfectly achieved over the finite time interval as the iteration step goes to infinity. Finally, three numerical examples are presented to demonstrate the effectiveness and wide application scope of the proposed control strategy. Keywords: Event-triggered control, fractional-order multi-agent systems, iterative learning control, non-identical fractional orders, output consensus tracking.
1.
INTRODUCTION
With the research of the integer-order multi-agent systems (IOMASs) [1–3] and the development of the fractional calculus [4], the fractional-order multi-agent systems (FOMASs) have been paid more and more attention. It has been found that the dynamics of multi-agent systems working in fluids or porous media should be described by the fractional-order calculus [5]. The integerorder systems can be considered as the special cases of fractional-order systems, though, due to the essential difference between them, most results of IOMASs can’t be simply extend to the FOMASs [6]. Thus, it is meaningful to investigate the FOMASs. A critical issue coming from the control of FOMASs is the consensus control problem, the goal of which is to design the distributed control protocol such that the states or outputs of all agents reach an agreement on the shared information [7]. The consensus problem of FOMASs was first considered in [8]. Afterwards, various kinds of models of FOMASs, such as heterogeneities [9], uncertainties [10, 11], and time delays [12] were proposed. Meanwhile, the consensus of FOMAS was studied by using different methods including feedback control [13], adaptive control [14] and slidingmode control [15].
All of the mentioned literatures are concerned with the time-triggered control strategy. Since the control input is updated continuously and each agent communicates with its neighbors at each instant, the time-triggered control strategy will consume large quantity of energy and lead to the communication resource overload. On the other hand, in the practical multi-agent systems, agents are often equipped with microprocessors, and the information is transmi
Data Loading...
