Multitask automatic manoeuvring systems using recursive optimal control algorithms

This paper presents a new approach to design a multitask automatic manoeuvring system by using recursive identification and optimal control algorithms. The shippsilas manoeuvring dynamics is described by an appropriate multivariable auto-regressive exogenous (MARX) model with unknown parameters. A recursive estimation algorithm is applied to estimate the time-varying parameters of the model. An optimal controller is used to calculate state feedback control gains and control signals. The automatic manoeuvring system consists of three subsystems: guidance, navigation and control. A shippsilas linear reference model, the waypoint, circle of acceptance and exponential decay techniques are used to generate a desired trajectory. The multitask integrated control system can do many tasks at seas such as maritime search and rescue mission, rudder-roll stabiliser, ship monitoring and information providing system for captain and pilot when berthing and unberthing and manoeuvrability index estimation system. Computer simulations to verify the multi-task automatic manoeuvring system are done for two vessels.