Research On Multi-UUV Attack And Defense Method Considering Target Threat Prediction

Unmanned underwater vehicles(UUV),as an important member of marine unmanned warfare mobile equipment,have the characteristics of good concealment and high intelligence,making UUVs widely used in military tasks such as reconnaissance,monitoring,and tracking.At the same time,UUV has played a huge role in safeguarding national maritime sovereignty.This paper takes the red and blue UUV offensive and defensive confrontation as the research background,with the red UUV cluster defense base as the research content,and conducts research on UUV trajectory prediction,threat assessment,intent estimation,and confrontation strategies:First of all,in order to better meet the actual requirements of offensive and defensive confrontation,UUV kinematics,dynamics models and UUV detection models are built.This paper simplified the model to get the UUV horizontal plane model,and designed a simulation experiment to verify the UUV model.At the same time,in order to control the UUV motion,the speed controller and the heading controller are designed through PID control principles,and the speed change and direction change tests are designed to verify the effectiveness of the controller;A single UUV system structure and a UUV cluster system structure are constructed for the needs of offensive and defensive confrontation.Secondly,based on the UUV mathematical model for trajectory feature data collection,the UUV trajectory prediction method is studied.Designed a gated recurrent neural network trajectory prediction model and a backpropagation neural network trajectory prediction model;then designed an adaptive parameter adjustment algorithm to optimize the hyperparameters of the gated recurrent neural network trajectory prediction model to improve the prediction accuracy,and finally verified by simulation comparison Its advantages.Thirdly,the multi-UUV threat assessment method is studied.The threat indicators of UUV are analyzed,a complete and reasonable threat assessment system is constructed,and a threat assessment method based on analytic hierarchy process is proposed.Aiming at the shortcoming of analytic hierarchy process that requires artificial determination of the judgment matrix,which leads to strong subjectivity,a method based on nuclear principal component analysis to rank UUV threats is proposed,and the selection of kernel function is studied in depth,and a hybrid kernel function is constructed.In order to better improve the extraction ability of the hybrid kernel function,the particle swarm algorithm is used to optimize its parameters,which proves its advantages compared with the analytic hierarchy process;In order to quickly judge UUV intentions,a single blue UUV intention estimation method based on fuzzy inference algorithm is proposed,and the effectiveness of the designed method is verified through UUV blue-side simulation cases.Finally,according to the tasks of the red and blue parties and the maneuverability of their respective UUVs,the confrontation strategy and task assignments of the UUVs of the red and blue parties are studied.According to the mission requirements of the red party to defend the base,a coordinated siege strategy,tracking strategy,and interception strategy were designed;for the proposed confrontation strategy,simulation verification was carried out under the UUV motion model and UUV motion controller;According to the threat assessment and intent estimation results Formulate a red party task allocation plan;Design a comprehensive simulation experiment to comprehensively simulate and verify trajectory prediction methods,threat assessment methods,intention estimation methods,and confrontation strategy.