Event-triggered-based Consensus Tracking Control Algorithms For Flexible Robotic Manipulators Multi-agent Systems

In recent years,with the rapid development of artificial intelligence and automation technology,the robotic manipulators multi-agent systems have been widely applied in the fields of medical treatment,industrial production,military and aerospace.Therefore,the control research of robotic manipulators multi-agent systems have attracted the attention of many scholars at home and abroad.Due to the complex system model and the interaction between flexible joints,it is difficult to obtain ideal results in the execution of tasks,which also makes the analysis,design and control of the robotic manipulators system research more challenging.The flexible robotic manipulators multi-agent systems are a whole system composed of multiple flexible robotic manipulators through communication with each other.Because of the unknown nonlinear characteristics of the robotic manipulators and various sudden situations in practical applications,it is a difficult research topic to realize the consensus tracking control of the flexible robotic manipulators multi-agent systems.In addition,the robotic manipulators multi-agent systems include single-leader multi-agent systems and multi-leader multi-agent systems,the control methods to be adopted and the control effects to be achieved are also different with the different number of leaders.Based on the above discussion,this thesis proposes an adaptive consensus tracking control algorithm based on event-triggered mechanism of relative threshold and an adaptive asymptotic tracking containment control algorithm based on event-triggered mechanism of switching thresholds for the flexible robotic manipulators multi-agent systems by using backstepping design method and adaptive radial basis function neural networks(RBFNNs)technology.The specific work is as follows:(1)For the flexible robotic manipulators multi-agent systems with single leader,an adaptive consensus tracking control algorithm is proposed based on command filtered method and event-triggered strategy of relative thresholds.In the first place,compared with most theoretical systems,the flexible robotic manipulators multi-agent systems with more practical significance are considered in this thesis.In the second place,the event-triggered strategy with relative thresholds mechanism is considered for the flexible robotic manipulators multi-agent systems,which greatly reduces the transmission frequency of the communication channels and saves the communication resources.Furthermore,the command filter technology is introduced into the frame of the traditional backstepping design,which makes the problem of “complexity of explosion” is well solved.In addition,the RBFNNs are used to model the unknown nonlinear functions,which overcome the design difficulties of the controller caused by the unknown nonlinearities in the system.Finally,the relative thresholds controller is constructed for each agent by sampling data from the original controller,which can ensure that all the signals in the closed-loop system are uniformly ultimately bounded(UUB),and the tracking errors between the followers and the leader can converge to a small neighborhood of the origin without Zeno phenomenon.In the end,the simulation results further demonstrate the availabilities of the proposed control schemes.(2)For the flexible robotic manipulators multi-agent systems with multiple leaders,an adaptive asymptotic tracking containment control algorithm is proposed based on a novel event-triggered mechanism with switching thresholds.Firstly,a two-step coordinate transformation method is proposed by fusing command filtered technique and backstepping method to avoid the repeated derivation of virtual control,which greatly reducing the computational communication burden and effectively avoiding the problem of “explosion of complexity”.Secondly,it is particularly difficult and challenging of the design controller due to the unknown nonlinear functions existing in the actual flexible robotic manipulators multi-agent systems.In order to obtain the desired controller,the RBFNNs intelligent estimation technology is introduced to deal with the design difficulties caused by the unknown nonlinearities in the system.In addition,by designing a novel event-triggered strategy with switching thresholds mechanism,the fixed threshold mechanism or relative thresholds mechanism corresponding to switching is considered respectively according to the input signal size,which greatly reduces the event triggering frequency,saves network bandwidth communication resources and avoids channel congestion.Finally,by skillfully applying mathematical processing methods and the classical Barbalat lemma,the proposed novel control protocol can not only ensures that all variables in the closed-loop system are UUB,but also the output of the followers can converge to the dynamic convex hull spanned by the output of the dynamic leaders,and the tracking errors between the followers and the leaders can asymptotically converge to zero.In the end,the simulation results verify the effectiveness of the proposed control schemes.In this thesis,the problem of consensus tracking control of flexible robotic manipulators multi-agent systems have been studied deeply.However,due to the complexity of the model structure studied,there are still many problems that need to be further solved,such as the bipartite consensus tracking control of flexible robotic manipulators multi-agent systems,security control under network attacks,etc.