復雜條件下多智能體系統(tǒng)的魯棒一致性控制研究
發(fā)布時間：2018-07-22 21:22
【摘要】：多智能體系統(tǒng)分布式協(xié)調控制是通過智能體之間相互作用來實現單個智能體疊加無法達到的整體功能,是當前系統(tǒng)控制領域的研究熱點和前沿。多智能體系統(tǒng)分布式協(xié)調控制的一個基本問題就是多種智能體系統(tǒng)的一致性控制問題。本論文針對復雜條件下多智能體系統(tǒng)的魯棒一致性控制問題,分析了時滯、外部擾動和參數不確定性對多智能體系統(tǒng)一致性的影響,設計了合適的控制算法使多智能體系統(tǒng)實現一致性并滿足期望的魯棒H_∞性能指標,并將其控制算法應用于無人機群的編隊控制。本論文主要研究內容和貢獻如下:針對具有固定時滯、外部擾動和參數不確定性的二階多智能體系統(tǒng)在二維平面上的魯棒旋轉一致性控制問題,提出了實現魯棒旋轉一致性所需要滿足的充分條件。通過系統(tǒng)模型變換,將多智能體系統(tǒng)的一致狀態(tài)與非一致狀態(tài)分離,得到降階系統(tǒng),從而將多智能體系統(tǒng)魯棒旋轉一致性控制問題轉換成降階系統(tǒng)魯棒H_∞控制問題;再基于該降階系統(tǒng),利用魯棒H_∞控制理論,推導出此類二階多智能體系統(tǒng)在二維平面上實現魯棒旋轉一致性所需要滿足的充分條件。針對具有時變時滯、外部擾動和參數不確定性的二階多智能體系統(tǒng)在三維空間中的魯棒旋轉一致性控制問題,提出了一種較低計算復雜度的魯棒H_∞旋轉一致性控制算法,提出了一種提高擾動抑制能力和時滯上限的控制參數優(yōu)化算法。利用變量代換和矩陣理論工具,將二階多智能體系統(tǒng)模型進行分解和解耦,從而將多智能體系統(tǒng)魯棒旋轉一致性控制問題轉換成解耦子系統(tǒng)魯棒H_∞控制問題;基于此解耦子系統(tǒng),利用Lyapunov-Krasovskii泛函方法和線性矩陣不等式的凸性,推導出了較低復雜度的時滯相關魯棒旋轉一致性的充分條件;基于該條件,利用一種圓錐補線性化算法和一種迭代算法,提出了一種時滯相關魯棒H_∞旋轉一致性的控制算法;拓展此控制算法,提出了一種提高擾動抑制能力和時滯上限的控制參數優(yōu)化算法。針對具有雙時變時滯、外部擾動和參數不確定性的高階多智能體系統(tǒng)的魯棒一致性控制問題,推導出了一個更低保守性的時滯相關魯棒一致性條件,提出了一種較低計算復雜度的魯棒H_∞一致性控制算法,提出了提高擾動抑制能力和時滯上限的控制參數優(yōu)化算法。利用變量代換和矩陣理論工具,將高階多智能體系統(tǒng)模型進行分解和解耦,從而將高階多智能體系統(tǒng)魯棒一致性控制問題轉換成解耦子系統(tǒng)魯棒H_∞控制問題;針對具有非一致自身狀態(tài)和輸入時滯的雙時變時滯情況,利用Lyapunov-Krasovskii泛函、Jensen不等式和帶自由權矩陣的Newton-Leibniz公式等工具推導出更低保守性時滯相關魯棒一致性條件;基于此條件,提出了一種時滯相關魯棒H_∞一致性的控制算法;拓展此控制算法,提出了一種提高擾動抑制能力和時滯上限的控制參數優(yōu)化算法�；诒疚年P于多智能體系統(tǒng)魯棒一致性控制的理論成果,提出了無人機群實現魯棒時變編隊的控制器綜合規(guī)則和算法。首先,通過一種反饋線性化方法,將無人機的非線性模型轉化成為一個雙積分線性化模型;其次,設計了基于局部無人機狀態(tài)信息反饋的分布式控制協(xié)議;再次,通過變量替換,將無人機群魯棒時變編隊控制問題轉化為多智能體系統(tǒng)魯棒一致性控制問題;最后,利用前文所獲得的關于多智能體系統(tǒng)魯棒一致性控制的理論成果,提出了無人機群實現魯棒時變編隊的控制器綜合規(guī)則和算法。
[Abstract]:Distributed coordination control of multi-agent system is a whole function which can not be achieved by the superposition of individual agents through the interaction between agents. It is a hot and frontier in the field of current system control. A basic problem of distributed coordination control of multi-agent system is the consistency control of multiple agent systems. Aiming at the problem of Robust Consistency Control of multiagent systems under complex conditions, this paper analyzes the influence of time delay, external disturbance and parameter uncertainty on the unity of multi intelligence system. A suitable control algorithm is designed to make the multi-agent system achieve consistency and satisfy the expected robust H_ infinity performance index, and its control algorithm is applied. The main research contents and contributions of this paper are as follows: for the robust rotation consistency control problem of two order multiagent systems with fixed time delay, external disturbance and parameter uncertainty in two-dimensional plane, the sufficient conditions to be satisfied for the realization of robust rotation consistency are proposed. The model is transformed into a reduced order system by separating the uniform state from the non consistent state of the multi-agent system. The robust rotation consistency control problem of the multi-agent system is converted to the robust H_ control problem of the reduced order system. Based on the reduced order system, the two order multi-agent system is derived by using the robust H_ infinity control theory. A robust rotational conformance control problem for two order multiagent systems with time-varying delay, external disturbances and parameter uncertainties is proposed for robust rotation consistency control of two order multiagent systems with time-varying delay, external disturbance and parameter uncertainty. A robust H_ infinity rotation consistency control algorithm with low computational complexity is proposed. A control parameter optimization algorithm which improves the ability of disturbance rejection and the upper limit of time delay is used to decompose and decouple the two order multi-agent system model by variable substitution and matrix theory, and then transform the robust rotation consistency control problem into the robust H_ infinity control problem of the decoupling subsystem; based on this decoupling subsystem By using the Lyapunov-Krasovskii functional method and the convexity of linear matrix inequalities, a sufficient condition for the robust rotation consistency of time delay related to low complexity is derived. Based on this condition, a time dependent robust H_ infinity rotation consistency control algorithm is proposed by using a conic complement linearization algorithm and an iterative algorithm. A control parameter optimization algorithm is proposed to improve the ability of disturbance rejection and the upper limit of time delay. For the robust conformance control problem of a high order multiagent system with dual time-varying delay, external disturbance and parameter uncertainty, a lower conservative delay dependent robust consistency condition is derived. A robust H_ infinity conformance control algorithm for lower computational complexity is proposed. A control parameter optimization algorithm is proposed to improve the ability of disturbance rejection and the upper limit of time delay. Using variable substitution and matrix theory, the model of high order multi-agent system is decomposed and decoupled, and the robust consistency control problem of the high order multi-agent system is solved. The robust H_ infinity control problem of the decoupling subsystem is transformed. For the case of two time time-delay with nonuniform self state and input delay, the Lyapunov-Krasovskii functional, the Jensen inequality and the Newton-Leibniz formula with the free weight matrix are used to derive the delay correlation robust consistency conditions for lower conservatism. A control algorithm for time-delay dependent robust H_ infinity consistency is presented, and a control parameter optimization algorithm is proposed to improve the disturbance rejection capability and the upper limit of time delay. Based on the theoretical results of Robust Consistency Control of the multi-agent system, a robust time-varying formation controller for the unmanned aerial vehicle group is proposed. In the first place, a feedback linearization method is used to transform the nonlinear model of unmanned aerial vehicle into a double product line model. Secondly, a distributed control protocol based on the information feedback of local UAV state information is designed. Again, the robust time-varying formation control problem of unmanned aerial vehicle group is converted to intelligence by variable substitution. The robust conformance control problem of the energy system is solved. Finally, using the theoretical results on Robust Consistency Control of the multi-agent system, the integrated rules and algorithms of the robust time variant formation of the unmanned aerial vehicle group are proposed.
【學位授予單位】：電子科技大學
【學位級別】：博士
【學位授予年份】：2017
【分類號】：TP13;TP18
本文編號：2138613
[Abstract]:Distributed coordination control of multi-agent system is a whole function which can not be achieved by the superposition of individual agents through the interaction between agents. It is a hot and frontier in the field of current system control. A basic problem of distributed coordination control of multi-agent system is the consistency control of multiple agent systems. Aiming at the problem of Robust Consistency Control of multiagent systems under complex conditions, this paper analyzes the influence of time delay, external disturbance and parameter uncertainty on the unity of multi intelligence system. A suitable control algorithm is designed to make the multi-agent system achieve consistency and satisfy the expected robust H_ infinity performance index, and its control algorithm is applied. The main research contents and contributions of this paper are as follows: for the robust rotation consistency control problem of two order multiagent systems with fixed time delay, external disturbance and parameter uncertainty in two-dimensional plane, the sufficient conditions to be satisfied for the realization of robust rotation consistency are proposed. The model is transformed into a reduced order system by separating the uniform state from the non consistent state of the multi-agent system. The robust rotation consistency control problem of the multi-agent system is converted to the robust H_ control problem of the reduced order system. Based on the reduced order system, the two order multi-agent system is derived by using the robust H_ infinity control theory. A robust rotational conformance control problem for two order multiagent systems with time-varying delay, external disturbances and parameter uncertainties is proposed for robust rotation consistency control of two order multiagent systems with time-varying delay, external disturbance and parameter uncertainty. A robust H_ infinity rotation consistency control algorithm with low computational complexity is proposed. A control parameter optimization algorithm which improves the ability of disturbance rejection and the upper limit of time delay is used to decompose and decouple the two order multi-agent system model by variable substitution and matrix theory, and then transform the robust rotation consistency control problem into the robust H_ infinity control problem of the decoupling subsystem; based on this decoupling subsystem By using the Lyapunov-Krasovskii functional method and the convexity of linear matrix inequalities, a sufficient condition for the robust rotation consistency of time delay related to low complexity is derived. Based on this condition, a time dependent robust H_ infinity rotation consistency control algorithm is proposed by using a conic complement linearization algorithm and an iterative algorithm. A control parameter optimization algorithm is proposed to improve the ability of disturbance rejection and the upper limit of time delay. For the robust conformance control problem of a high order multiagent system with dual time-varying delay, external disturbance and parameter uncertainty, a lower conservative delay dependent robust consistency condition is derived. A robust H_ infinity conformance control algorithm for lower computational complexity is proposed. A control parameter optimization algorithm is proposed to improve the ability of disturbance rejection and the upper limit of time delay. Using variable substitution and matrix theory, the model of high order multi-agent system is decomposed and decoupled, and the robust consistency control problem of the high order multi-agent system is solved. The robust H_ infinity control problem of the decoupling subsystem is transformed. For the case of two time time-delay with nonuniform self state and input delay, the Lyapunov-Krasovskii functional, the Jensen inequality and the Newton-Leibniz formula with the free weight matrix are used to derive the delay correlation robust consistency conditions for lower conservatism. A control algorithm for time-delay dependent robust H_ infinity consistency is presented, and a control parameter optimization algorithm is proposed to improve the disturbance rejection capability and the upper limit of time delay. Based on the theoretical results of Robust Consistency Control of the multi-agent system, a robust time-varying formation controller for the unmanned aerial vehicle group is proposed. In the first place, a feedback linearization method is used to transform the nonlinear model of unmanned aerial vehicle into a double product line model. Secondly, a distributed control protocol based on the information feedback of local UAV state information is designed. Again, the robust time-varying formation control problem of unmanned aerial vehicle group is converted to intelligence by variable substitution. The robust conformance control problem of the energy system is solved. Finally, using the theoretical results on Robust Consistency Control of the multi-agent system, the integrated rules and algorithms of the robust time variant formation of the unmanned aerial vehicle group are proposed.
【學位授予單位】：電子科技大學
【學位級別】：博士
【學位授予年份】：2017
【分類號】：TP13;TP18
【參考文獻】
相關期刊論文 前1條
1 ;ROBUST CONSENSUS AND SOFT CONTROL OF MULTI-AGENT SYSTEMS WITH NOISES[J];Journal of Systems Science and Complexity;2008年03期
，本文編號：2138613
本文鏈接：http://sikaile.net/kejilunwen/zidonghuakongzhilunwen/2138613.html
