Model Free Adaptive Formation Control Method And Its Application Of NAO Robot

Multi-agent system is an organic combination of multiple agents according to certain methods or rules.Compared with single agent system,a multi-agent system not only has the characteristics and advantages of all single agents,but also can accomplish more complex and difficult tasks through cooperation among agents.Multi-agent formation control is the basis of multi-agent cooperative motion control,and is also one of the hot research directions.At present,multi-agent formation control methods need homogeneous systems and known models.Model-free adaptive control is a data-driven control that requires no modeling and uses only system input and output data to design the controller.Model-free adaptive control provides a new idea for unknown heterogeneous multi-agent formation control.In this thesis,the formation control problem basing on heterogeneous nonlinear multi-agent systems is studied,and a model-free adaptive formation control scheme is proposed.The effectiveness of the proposed formation control scheme is verified by MATLAB formation simulation of wheeled inverted pendulum and NAO robot formation experiment.The specific work is as follows:(1)Several common formation control methods and model-free adaptive control algorithms are studied.The model-free adaptive control method based on full format dynamic linearization is combined with graph theory method.A full form dynamic linearization based model free adaptive formation control(FFDL-MFAFC)scheme for unknown heterogeneous nonlinear multi-agent systems is proposed.(2)The heterogeneous nonlinear wheeled inverted pendulum system uses FFDL-MFAFC scheme for formation simulation and comparative study based on MATLAB.The wheeled inverted pendulum system based on heterogeneous nonlinearity uses FFDL-MFAFC scheme for MATLAB formation simulation and comparative study.The simulation includes two parts.The first part is the speed tracking control of the wheeled inverted pendulum.The state feedback method and PID control method are used to complete the balance control and speed control of the inverted pendulum respectively.The second part is to use FFDL-MFAFC scheme for the formation control of three wheel inverted pendulum systems,and compare it with the PID formation control scheme.The results show that FFDL-MFAFC scheme has better control effect.(3)Using three NAO robots to build a formation control experimental platform,and carrying out the NAO robot formation straight walk experiment and curve walking experiment based on the platform.FFDL-MFAFC scheme and PID formation control scheme are used in each experiment respectively,and the control effects of the two schemes are compared.The results show that the control effect of FFDL-MFAFC scheme is better than that of PID formation control scheme.Simulation and experiment both verify the effectiveness and superiority of FFDL-MFAFC scheme.