Research On The Track Control Algorithm Of Unmanned Surface Vessel

With the change of the war way, unmanned combat system has become the main trend of the future. Unmanned Surface Vessel (USV) as one of the important branches, more and more nation pay attention to it. Among them, the autonomous navigation technology is the core of the USV’s control, mainly in the intelligent risk aversion and path tracking control these two areas. This article will focus on the USV’s track control algorithm. Ensure that the USV can sail according to the planning route, and can provide beneficial information for accurate strike and complete special combat mission for future wars.USV is a typical nonlinear, motion model uncertainty system, also suffer the wind, wave and current external environment interference while sailing on the sea, which makes the USV’s accurate path tracking is very difficult. Neural network’s powerful learning ability makes it become the research focus in many fields. In this paper, in order to improve the accuracy, robustness and stability of the USV’s tracking, we carry on an in-depth research and analysis on neural network control theory and its application in the USV’s course control and track control.This paper first introduces the indirect and direct type these two kinds of track control schemes, and studies the calculation method of the USV’s position, direction and tracking error. On the basis of USV’s space three degrees of freedom movement, establish its planar motion’s linear and nonlinear mathematical models, it also gives the wind, wave and current interference model. Then introduces the basic principle of BP neural network, For indirect track control, combining neural network and PID control, propose PID course controller which is based on BP neural network. With the simulation tools Simulink in MATLAB, do the simulation research of the traditional PID course control and the BP neural network PID course control respectively. Simulation results show that BP neural network PID course controller which implements controller parameters’on-line adjustment, In the wind, wave and current interference, its control effect is better than the traditional PID control obviously. Finally, propose the concept of neural network on-line direct track controller for direct track control, according to the direction of the compass, navigational speed, rudder angle, GPS positioning, electronic charts and other information, controlled by coordinates, build the simulation block diagram in Simulink. The BP neural network controller is realized by S function module. The simulation results show that neural network on-line track controller has a good path tracking accuracy.