| The accurate recognition of mobile robot for environment information is the basis of autonomous navigation and intelligent control. In this context, this paper takes the pattern recognition of mobile robot as the research focus. Through the research of the obstacle pattern recognition algorithm, an improved neural network pattern recognition method is proposed in this paper. The proposed method is applied to the pattern recognition of obstacles around the mobile robot, and it can realize the accurate classification of the obstacle mode, which lays the foundation for the next decision of the mobile robot.This paper firstly summarizes the development situation of mobile robot and pattern recognition. Through the comparison of the obstacle detection technology, the scheme of ultrasonic detection is confirmed. Then put forward a kind of improved evolutionary firefly algorithm. We applied the improved firefly algorithm to the optimization of BP network which increases the training speed and accuracy of the network. Finally, through the analysis of the mobile robot and the working conditions,we designed the obstacle pattern recognition system and applied the improved neural network patter recognition methods in mobile robot system. Through the system test and analysis, we achieved the desired results, improved the mobile robot' cognitive ability to the environment. |
PDF Full Text Request