Research On Three-axis Turntable Servo System Based On Model Predictive Control Algorithm

In order to shorten the study period of underwater vehicle,reduce the cost of experiment,and improve the security and reliability of the system,the requirements of underwater vehicle simulation platform are becoming higher and higher.In the traditional semi-physical simulation platform of the underwater vehicle,the sensor and the attitude display of the underwater vehicle are simulated by the virtual sensor and the three-dimensional scene of the host computer,respectively,but the virtual sensor is based on the calculated output of the control system,Which is a great error between the experimental data obtained from the real sensor of the underwater vehicle and the velocity and acceleration of the underwater vehicle's attitude calculated by the dynamic model.3D simulation of the aircraft,and can not carry out the real attitude simulation of underwater vehicle.This is not enough to establish the underwater vehicle simulation platform to achieve the software and hardware of the underwater vehicle control system and the real system software and hardware exactly the same intention.In order to achieve a realistic simulation of underwater vehicle attitude,a three-axis turntable servo control system is used to simulate the attitude of underwater vehicle in the attitude control system of a small underwater vehicle.In this thesis,the function extension and technical upgrading of the underwater vehicle semi-physical simulation system are based on the "Small Underwater Vehicle Motion Characteristic Test Instrument Project"(No.YZ201441)of the Chinese Academy of Sciences(CAS),The attitude simulation of the underwater vehicle with three-axis turntable is mainly studied.The structure and algorithm of the control system are designed and validated.In this thesis,the dynamic model of three-axis turntable is studied,and the three-axis turntable DC motor model and three-axis turntable controlled object model are analyzed.Based on the mathematic model of the three-axis turntable,the time-varying state space equation required for the Model Predictive Control(MPC)algorithm is established.The MPC optimal objective function is set and the optimal objective function.In addition,the dynamic model prediction in MPC control algorithm is simulated by neural network and support vector machine,and the particle swarm optimization algorithm is applied to the rolling optimization.Finally,a new improved model prediction algorithm is obtained.Besides,the influence of the controller parameters on the control effect is analyzed by a large number of experiments,and the controller parameters are determined.Emphasis on the improved MPC control algorithm in the three-axis turntable control system simulation tests,reflecting the improved new model-side control algorithm superiority.