Data‐Driven Cooperative LOS Guidance‐Based Constrained Path‐Following of Multiple Unmanned Surface Vehicles

ABSTRACT This paper investigates the model‐free coordinated path‐following problem for multiple input‐constrained unmanned surface vehicles (USVs), whereby a novel data‐driven cooperative LOS guidance‐based constrained path‐following (DCLOS‐CPF) scheme is innovatively developed. Unlike previous continuous cooperative guidance methodologies, the data‐driven control strategy is first deployed to deal with the complex path parameter subsystem, and a simple data model for constructing the relationship between the path parameter and the target surge speed is efficiently established. Then, based on the established data model and the cooperative path parameter tracking error, a data‐driven coordinated surge guidance law together with distributed LOS heading guidance scheme is thereafter devised to achieve accurate path‐following and stable coordination at the kinematic level. Afterward, to cope with the input saturation and unknown kinetic dynamics at the kinetic level, two efficient data‐driven constrained velocity controllers based on data‐driven anti‐windup compensation technology and data‐driven sliding‐mode control strategy are further developed. Finally, the stability of the entire closed‐loop DCLOS‐CPF control system is rigorously validated by virtue of the special cascaded guidance‐control structure and the input‐to‐state stability criteria, and the effectiveness and superiority of the proposed DCLOS‐CPF approach is testified via simulation studies.