Based on the model-free adaptive control (MFAC) theory, the heading control problem of an unmanned surface vehicle (USV) with uncertainties is researched. First, the compact form dynamic-linearization-based MFAC (CFDL-MFAC) method and its inherent failure problem with USV heading control are analyzed. Second, to solve the problem that the distinctive dynamic characteristics of USV heading control subsystem make the CFDL-MFAC unsuitable, the redefined output CFDL-MFAC (RO-CFDL- MFAC) method is proposed by introducing redefined output gain. Then, theoretical analysis shows that the RO-CFDL-MFAC method can be applied to the heading control subsystem, that is, it makes the heading control satisfy the quasi-linear assumption of MFAC. Next, the minimum range of redefined output gain and the main influences are analyzed, and the stability of RO-CFDL-MFAC method is theoretically proved. Finally, the simulation studies and field experiments demonstrate the validity and engineering practicability of the RO-CFDL-MFAC method.

Underwater terrain matching navigation technology is an important research area for the underwater navigation of Autonomous Underwater Vehicles (AUVs). Terrain matching navigation can realise long-term, subtle, all-weather, and high-precision underwater AUV navigation. In this paper, the research status of the application of AUV underwater terrain matching navigation is reviewed, the system composition, theory and terrain matching methods of underwater terrain matching navigation are summarised and the advantages of a multi-beam bathymetric system in underwater terrain matching navigation are discussed. The current research thoughts are summarised, the key issues are pointed out, and possible future development trends are discussed.