Research On Obstacle Avoidance And Control Of Navigation For Mobile Robot Based On The Multi-sensor Information Fusion Technology

With the development of the sensor technology,the application of the multisensor information fusion has become a very popular research area in the mobile robot. The technology of fusion provides a type of technology approach for the robot to probe the uncertain and unknown environment,and this is the foundation of high-quality intelligent behavior.Fuzzy logic and neural network are key methods for multisensor information fusion at present, but lack of fuzzy logic is self study and adaptive ability,lack of neural network is the expression of basic knowledge. In this paper,we used the fuzzy logic control algorithm and fuzzy neuron networks based on the T-S(Takagi-Sugeno) model in avoiding obstacles and navigation.The merit of neural network and fuzzy logic is merged into this algorithm,so the algorithm have the fuzzy reasoning ability of disposing the uncertain information and the study ability based on the sample data.First of all,we studied the technology of multisensor information fusion,focus on the fuzzy logic algorithm and fuzzy neuron networks based on the T-S model. Secondly,the paper presented sensors which were used in avoiding obstacles and navigaiton,selectively analyzed sensors' system of the robot. In the sensor system, ultrasonic sensor and infrared sensor supply the obstacle's information in unknown environment, electronic compass and GPS which constitute the positioning system supply the position of the robot and the target information.At last,we took the fuzzy logic and fuzzy neuron networks based on the T-S model in order to navigate and avoid obstacles,and did the simulation in MATLAB. Compared with the simulation result,the algorithm this paper presented is better than fusion algorithm based on the fuzzy logic in navigation and avoiding obstacles for the robot.