Research Of The Submarine Steering Neural Network Control System

With the development of the modern shipping industry and the demand of the modern sea war, the simple steering control system has already been replaced by the steering control system on which the modern control theory and the computer control system are applied. Quantitative analysis and studies on the steering control system being carried on, this paper designs the control system based on neural network PID algorithm and the self-adaptation.First, this paper introduces main components of the automatic steering, analyzes two coordinates commonly used in shipping and studies the rotation relationship between them. By analyzing the general equation of submarine motion space and the strength situation, this paper gives the submarine motion equation of six degree of freedom. For the convenience in study, we use the standard submarine space motion model directly, including the axial equation, the horizontal equation, vertical equation, inclining equation, vertical to incline equation and go off course equation. In order to meet the demands of the steering control system, this paper simplifies the general equation of the steering motion and analyzes, its stability.Secondly, this paper introduces the history and current development of the neural network, and provides classical increment type PID controls algorithms. After studying the composition and algorithm of the single neuron adaptive PID control system, this paper gives the deriving formulae and calculation procedure of the adaptive PID based on neural network, and the improved algorithm is also provided in this paper.Finally, the evaluation index and the dynamic quality analysis of the submarine steering control system are analyzed in the paper, including the steering system stability, its error analysis under interference, and the submarine static index analysis etc. With the neural network PID control algorithm being applied in the submarine steering control system, using K-T equation as the submarine movement model, and the neural network 4-5-3 being selected, adaptive PID control algorithms being used in adjusting the adaptive parameter, the neural network control system of the submarine steering is designed, the systematic simulation is carried on, and the artificial simulation curve and the result analysis are also provided in this paper.