| Multiple robots system has the characteristics of good spatial distribution,information complementation,low cost,and excellent performance.It can complete complex tasks which are difficult to be completed by a single robot.It can realize the task diversion of complex tasks,improve the ability of a single robot,and be flexibility,robustness and sociality.Cooperative target tracking control of multiple robots can be widely used in military,space exploration,environmental monitoring and other fields,and it has important practical application value.Especially,combining the inherent characteristics of robots such as nonholonomic constraints and nonlinearity as constraints with distributed cooperative motion control,the cooperative target tracking control of multiple robots has more important theoretical research significance.Combined with practical applications and analysis of existing research results,there are still several problems to be solved urgently in the multiple robot cooperative target tracking control:(1)Relatively simple robot models are often used.The cooperative control algorithm based on linear system is designed.Once the above theory is applied to the actual robot model,it is difficult to obtain the ideal control effect.(2)The designed robot application environment is too simple.Practical problems such as external interference,obstacles,and limited communication range are often not considered.If these problems are ignored,it will bring limitations to practical application.(3)The "uncertainty" problems of robots are often ignored.The high nonlinearity and complexity of robots will lead to modeling bias and unmodeled dynamics.Once these problems are ignored,the accuracy of cooperative control will affected.(4)The possible problems of robot individuals are often ignored.For example,the robot will go wrong and the output of the actuator is limited.If these problems are ignored,the performance of system will be reduced and it will be unstable.In summary,combined with the difficulties in practical applications and the deficiencies of previous research,Considered ground multi-robots satisfying non-holonomic constraints and underwater multi-robots and space multi-robots satisfying complete constraints,this paper takes multiple robots as the application object,and multiple nonholonomic constraints models and multiple Euler-Lagrange models are used,cooperative control related theory and nonlinear control theory are taken as the theoretical research basis.In the paper,two kinds of formation to complete the cooperative target surround tracking and cooperative target follow tracking are researched separately,and it mainly includes the following aspects:(1)In view of the lack of information caused by the local communication between the target and robot,a distributed estimation strategy that uses only local information to estimate the target position is proposed;A tracking differentiator with a filtering function is used to estimate the target speed to prevent noise interference caused by direct differentiation;on this basis,a distributed control law is proposed based on graph theory and backstepping method to realize cooperative target surround tracking control of multiple nonholonomic robots.Finally,the effectiveness of the proposed control law is verified by theoretical proof and simulation experiments.(2)Aiming at the limited perception and interaction ability of robots and robot fault-tolerant control,using the distance and azimuth information between the robot,the target,and the neighboring robot,a distributed control law independent of the number of robots is designed based on the vector field method.The control law is easy to expand,reduce and increase the number of robots in the system.Fault-tolerant control of cooperative target surround tracking for multiple nonholonomic robots is realized.Finally,the stability of the system is proved by Lyapunov's theorem and Barbalat's lemma,and the effectiveness of the control law is verified by simulation experiments..(3)The influence of the control effect caused by the inaccuracy of underwater robot and space robot modeling and the interference and obstacles in the environment,a robust inner-outer loop cooperative controller is designed based on the multiple Euler-Lagrange models.Null-space-based Behavioral Control(NSB Control)is adopted in the outer loop,which describes and maps different tasks through task-first motion control inverse kinematics,and realizes the cooperative target follow tracking with obstacle avoidance;regarding the problems of "slowly variable linearization uncertainty",the Adaptive Proportion Differentiation-Sliding Mode Control(APD-SMC)method based on adaptive control is designed in the inner loop to achieve a model-independent and robust control system.The theory proves the stability of the inner and outer loop controllers,and through comparative experiments with NSB+Sliding Mode Control(SMC),NSB+Adaptive Sliding Mode Control(ASMC)and NSB+PD,the superior control performance of the proposed controller is verified.Regarding the problems of "rapidly changing nonlinear uncertainties",Radical Basis Function Neural Networks Proportion Differentiation-Sliding Mode Control(RPD-SMC)method is designed in the inner loop to realize a strong learning ability,simple,model-free and strong robust control system;finally the robust control of multiple Euler-Lagrange robot cooperative target tracking is realized.(4)Aiming at the input constraint problem in practical applications,the RPD-SMC method is improved by using the bounded arctangent function,and a Bounded Radical Basis Function Neural Networks Proportion Differentiation-Sliding Mode Control(BRPD-SMC)method is proposed,which guarantees the bounded input and the advantages of RPD-SMC,and the bounded,model-free and robust control system is realized.Finally,a rigorous theoretical analysis for the stability of the system is carried out,and comparison experiments with NSB+ASMC and NSB+PD verify that the proposed controller has good control performance. |
PDF Full Text Request