Autonomous Knowledge Generation Mechanism Of Intelligent Transportation Robot Based On Language Function

With the development of advanced technologies such as artificial intelligence,Internet of Things and machine vision,robots are widely used in various fields.With the support of wireless communication and high-speed networks,intelligent traffic robots are expected to replace traffic police in the traffic field to complete traffic command tasks,relieve the work pressure of traffic police through "machine substitution" and "human-machine collaboration",enhance urban traffic management,reduce traffic accidents and help the intelligent and intelligent development of urban traffic.Apply robot technology to the field of traffic command,design intelligent traffic robots,let the robot have traffic command capabilities,based on speech recognition technology,add language functions to robots through natural language processing technology,and realize human-computer interaction based on natural language.Intelligent traffic robots can understand human natural language,can conduct effective traffic command according to traffic police voice instructions,and complete traffic command tasks.The main research content of this topic is as follows:(1)According to the needs of traffic command tasks,combined with the use scenarios of intelligent transportation robots,the scheme selection and structural design of the robot’s mechanical arm,fuselage structure and mobile platform are respectively carried out.According to the functional requirements of the intelligent transportation robot,the control scheme of the total system of the robot and the control scheme of the manipulator and the mobile platform are selected,and the dynamic model of the robot is simplified.The simulation experiments show that the overall stability of the robot is good.(2)Provide technical support for intelligent traffic robots to obtain natural language information based on speech recognition technology,collect and preprocess speech signals according to the principle of speech recognition,and write programs to carry out secondary development of i FLYTEK speech recognition API interface,through The cloud speech recognition service architecture builds a real-time speech recognition system,then processes the recognition results and outputs text information,and finally conducts experimental tests on the speech recognition system.(3)According to the types of text errors and common text error correction methods,construct a homonym list and a homonym confusion set,use the corpus to train the neural network model,and correct the errors in the sentences after speech recognition.Aiming at the problem of voice command recognition errors in traffic command tasks,a homonym confusion set of error-prone words was established,and a text error correction method combining error-prone word confusion sets and a binary language model was used to complete the text information correction of voice commands,and passed the experimental test It shows that the text error correction method has a better effect.(4)Use natural language processing technology to process the text corpus,choose the method based on the combination of edit distance and cosine similarity calculation to calculate the instruction text similarity,and return the text instruction with the highest similarity.Finally,the voice command is issued to the intelligent traffic robot,and the movement of the robot itself and the execution of the traffic command action are tested respectively.The feasibility of the traffic command task of the intelligent traffic robot based on the language function is verified through experiments.