3d Simulation System Of Micro Robotic Fish For Obstacle Avoidance Research

Micro-bionic robot fish has become a hot issue of robot fish research field because of their many advantages such as small size, high- driving performance, low noise, good flexibility and adaptability. In the course of the study, 3D simulation system using virtual reality technique plays an important role. It can reduce the risk, unpredictability and the costs during the research and development, short the development cycle.Taking the ICPF-driven micro-bionic robot fish as the prototype, a 3D simulation system of micro biomimetic robot fish for obstacles avoidance is developed in this thesis.Firstly, the approach of the static modeling of robot fish is studied through OpenGL with 3DMax modeling tools and PolyTrans 3D file format conversion software. Based on the analysis of the fish tail swinging characteristics, the discrete equation of robot fish swinging is obtained, thus the dynamic model of robot fish is achieved. Real simulation environment is established using OpenGL rendering function such as clearing screen and fogging.Secondly, vivid and feature-rich man-machine interface is realized, which not only provides the functions of setting simulation environment, controlling system and multi-angle observation, but also meets different needs of users.Thirdly, a virtual ray method for simulating the infrared sensor to detect obstacles is proposed. The difficulty of coordinate transformation calculation is simplified by selecting robot fish and sensor local coordinates which are the same with world coordinates of 3D simulation system. Based on bounding sphere method to detect sphere type of obstacles and bounding box method to detect cylinder type of obstacles are proposed, which solves the problem of the difficulty of detecting different types of obstacles in simulation environment.The composite fuzzy controller which contains two angle fuzzy controllers and a velocity fuzzy controller is designed, which can decide robot fish's obstacle avoidance behavior. The realtime of robot fish obstacle avoidance is met because this kind of composite fuzzy controller needn't establish the environment model and there isn't complicated calculation.Finally, the experiment of robot fish obstacle avoidance is done in various obstacles environment of the 3D simulation system. The results demonstrate that the 3D simulation system developed in the thesis could more vividly and reliably simulate the real situation, the proposed virtual ray method for simulating the characteristic of sensor could detect all kinds of obstacles in the environment, the designed composite fuzzy controller could effectively direct robot fish to safely arrive at the target from the start and the system simulation environment is vivid and the interface is user-friendly.