Robust Neural Network and Its Application to Course-keeping for Ships

A feed-forward network with seven layers was designed and trained to acquire an open loop controller of course-keeping for ships. The NNC controller with five inputs and one output could follow the tacks of typical course curve with good precision, then a closed loop control system was constructed using closed loop gain shaping algorithm. The robustness of the system is improved because the control strategy connects the adaptability and nonlinear mapping of neural network with robustness of closed loop gain shaping algorithm. This method has the advantages of simple design procedure and obvious physical sense