The system of the zero-speed fin stabilizers controlled by common PID controller can hardly obtain satisfying performance,so an improved adaptive single neuron controller is developed,and the capability for nonlinear processing is improved without increasing the structural complexity of the neuron.To reduce the requirement of the power for driving servo system,from the viewpoint of the quadratic performance index employed in optimal control,the indirect constraint to the driving power is realized by means of online updating parameters in the neuron adaptively.The Lyapunov stability of the system with this control method is proved.Compared with the common PID control under different sea conditions,the advantage for controlling zero-speed fin stabilizer system with nonlinear input is demonstrated.