| Underwater vehicles are the main tools for helping mankind to realize,develop and utilize ocean.Realize information exchange and independent collaborative work among multi-underwater vehicles will be a hot topic in the future,which can overcome the drawbacks for manned submersibles in short work time and small operating rang.In this thesis,the design of experiment platform system for underwater vehicle with the function of control,communication and positioning is provided,and the positioning accuracy and path planning of underwater vehicle in the experimental system are studied.The main research works are given as follows:Firstly,the experimental platform system for underwater vehicle with wireless communication function is built.The system mainly includes the following two parts:(1)The design of small AUV underwater vehicle is based on STM32F4 core board assisted wireless communication module,depth meter,attitude sensor and three propellers;(2)The design of Visual Global Positioning Algorithm is based on CamShift Target Tracking.The experimental platform system has some basic functions such as horizontal visualization based on global vision,attitude acquisition,moving trace display,and information interaction between underwater vehicle and host computer.Secondly,considering the stability and precision of the posture angle in the complex underwater environment,the attitude calculation method of the multi-sensor fusion for underwater vehicles is given.The magnetometerHMC5883 L and inertial sensor MPU6050 can be used as the calculation of the original algorithm design,and two sets of complementary quaternion are obtained by the use of gradient descent method and quaternion Longge-Kutta method.The adaptive Kalman filter is furtherly applied to correct the quaternion of the magnetometer and the accelerometer and the deviation of the gyroscope.The observed noise is estimated and adjusted by two fuzzy control blocks such that a high accurate and stable posture angle can be obtained easily.Finally,aiming at the problem of slow initial convergence and local optimum for colony algorithm in path planning,an improved ant colony algorithm is proposed.Above all,the artificial potential field and the heuristic information are used to improve the randomness and blindness of ant colony algorithm in the initial stage.Then,the ant colony algorithm global search mechanism and excellent positive feedback effect are applied to improve the convergence of the algorithm such that the underwater vehicles are guided to realize the optimal path planning rapidly.Finally,the effectiveness of the experimental platform and the proposed path planning algorithm are verified. |
PDF Full Text Request