The Research On Simulation Modeling And Parameter Identity For Submarine

Because using the simulator training can train the steering more quickly and accurately,and can reduce a lot of spending, the simulator training obtains the widespread support and application in various industries, especially the military branch. The sailing training simulator provides a realistic simulation platform for submarine manipulation control personnel, and can carry out various submarine manipulations training accordingly. In order to train the navy better, the accuracy and practicability of the motion model is very important. Given the large parameters of the general simulation model, the complexity of the calculation and the error between the real traffic data, a proposal to simplify the model based on the sensitivity and a method using ABC(Artificial Bee Colony) to optimize the coefficients is put forward on the premise of guaranteeing the simulation quality in this paper which improve the manipulation training skillfully.Firstly, the general, classical model is given in the paper, and the simulation and control model is provided base on the conventional assumptions. Besides, the general calculation of motion model with the six degrees is derived, which lays the foundation for subsequent simplified model and parameter optimization.Then, zero sensitivity index is put forward based on the sensitivity index of coefficients.The minimum model is deduced based on the sensitivity. The simulation comparison between control model and minimum model is made in this paper. In addition, the influence of hydrodynamic coefficients on maneuvering state is tested under different initial conditions,and then simplified models under varied conditions are retrieved from the coefficients varied rules.Finally, because of the data differences between the simulator results and the real submarine,ABC is proposed to derive the training model. The deviation of the data between minimum maneuvering model and real maneuvering test is used as the fitness function of the ABC. The expression of fitness function, the number and the range of the parameters are determined in the dissertation. The standard algorithm is improved and the results show its validity.Aiming at the study of model simplification and parameter identification, the proposed scheme is simple,easy to operate,and has strong universality which fits for the similar study.The proposed scheme increases the availability and practicability of the simulator, saves a large amount of computer resources, and improves the soldiers’ practice level with the practical significance.