Study On GA-NN Stabilization Control Of Strap-Down Mine-Hunting Sonar Array

Although there are many kinds of means of anti-mine, the mine-hunting, which has many merits such as celerity, go-aheadism and security is regarded as the most important means of anti-mine by many national navies. The carrier strap-down sonar array is the base of the sonar array; its attitude control will determine the success of the mine hunting.In the control system mathematic platform of the strap-down mine-hunting sonar array (SMSA), considering the line displacement caused by the hunting ship swaying and the disconnected point caused by the trigonometric function transformation, this paper proposes the correctional mathematic platform of control system for strap-down mine hunting sonar array (CS-SMSA) in which corrected order attitude (ψ_t) and corrected order pitching (β_l), and use them replace order attitude ψ_g and order pitching β_g which used in the old mathematics platform. Deduced corrected control system mathematic platform of SMSA, it resolves the line displacement the disconnected point successfully.Because the genetic algorithms (GA) search speed is slow, this paper proposes an improved GA (IGA), and presents a diagonal recurrent neural network (DRNN) based on IGA. η_l,η_D,η_O (learning rate of the input layer, hidden layer and output layer) and q (the number of hidden layer) of DRNN optimized by GA. The simulation results show that the DRNN obtained by GA optimization have good quality.The system of three frames moving for the SMSA is a close coupling system. This paper proposes to decouple CS-SMSA with the neural network, and establishes the distributed decoupler based on the DRNN, and demonstrated the training algorithm for the neural network decoupling of CS-SMSA. This decoupler compensates the coupling influence of three frames moving for CS-SMSA.With the application of DRNN and GA-DRNN on the mine-hunting sonar array servo system, we establish servo systems for SMSA attitude, pitching and rolling. These servo system based on DRNN and GA-DRNN gain lower tracking errors, shorter settling time and a stronger robustness.This paper study on SMSA three freedoms primary-slave coordinating-control...