Design And Research Of Auv Control System Based On Vision

Underwater exploration has become a new strategy for the development of all countries,and the underwater vehicle has become the focus of research for the relevant institutions as one of the important means to explore the oceans and maintain the marine equipment.As the sonar and other sensors in the close operation is limited,visual-based underwater vehicle motion control has been rapidly developed.In this thesis,the visual servo control of underwater vehicle was studied from three aspects: the design of control system,the treatment of underwater refraction distortion and the realization of control method.In this thesis,a set of hardware platform for underwater vehicle control system was designed and built,including controllers,sensors,drivers and power management system.We got the sensor data through the I2 C and UART.The data acquisition process of camera as position sensor was described in detail.The centralized Kalman filter method was used to fuse the sensor data.Experiments showed that the fusion algorithm can improve the reliability and accuracy of the data.The problem of underwater visual pose measurement is that the multilayer medium causes the light to refract to cause serious distortion of the picture,which will reduce the accuracy of the visual measurement.In this thesis,the geometrical model and mathematical model of underwater refraction distortion recovery were established according to the imaging system structure.The correlation curve was drawn by Matlab and the imaging law was obtained.The image distortion recovery algorithm was developed according to the imaging rule and the pose estimation algorithm.Based on the underwater measurement of the classical distortion recovery model and the method described in this thesis,it is concluded that the measurement accuracy of the method described in this thesis is within the allowable range,and the measurement accuracy is better than classic distortion recovery model under the small angle and large displacement.For the motion control system of the vehicle,this thesis first established the coordinate system of the vehicle,and then discussed the dynamic model of the vehicle.The relationship between the motion parameters of the vehicle in different coordinate systems was analyzed.We discussed the mission planning and this thesis mainly studied the hovering control of the vehicle.The hovering control of the vehicle was divided into three parts: searching target,attitude adjustment,stable hover.Multi-PID controller control strategy was used to stable hover,and for a small amount of posture changes we added the dead zone and pulse control.Vehicle suspension control platform was built and hover accuracy and stability was evaluated by the visual measurement method.The experimental results showed that the control system designed in this thesis can meet the requirements of hover accuracy and stability.