A Study On Mechanic ARM Voice Control System

Robots with multi-function have been widely used in all walks of modern society.However,the robot's intelligent development is limited due to the simple traditional control mode and operational inconvenience.How to improve the robot's human-computer interaction ability and to find out a more convenient way of human-computer interaction become significant branches of robot research.As language is the most important way of human communication,using voice command to control the robot to complete the corresponding action will make human-machine interaction more convenient and efficient.This study,based on the analysis of the hardware and software platform,researching on the voice control system of manipulator which relying on the background of the application of the logistics sorting process.Aiming at the problem that the interference of noise affects the accuracy of speech recognition,an improved spectral subtraction is proposed to denoise the collected speech instruction,so as to improve the voice recognition accuracy of voice command in the robot voice control system.This paper mainly studies the following aspects:Firstly,the model structure of multi-path logistics sorting system based on machine hearing is established,based on the analysis of hardware and software of six-degree-of-freedom MOTOMAN mechanical arm experimental platform.Furthermore,according to the different requirements of the logistics sorting system,the subsystems of each part are designed and extended in the control mode.Finally,the development ideas of invoking the Motocom32 dynamic link library to grab and sort of different types of goods are determined.Secondly,Aiming at the problem that the interference of noise affects the accuracy of speech recognition,an improved spectral subtraction is proposed to reduce the noise of the system.In this algorithm,the power spectrum of speech signal is estimated according to the theory of multi window spectrum estimation.Based on power spectral density of smooth processing to improve the existing algorithms.Compared with the simulation,the improved algorithm can separate the pure speech more effectively,and get the voice signal with higher signal to noise ratio.In order to overcome the weakness of the traditional hidden markov model,the hidden markov model and neural network model are fused,and the mixed model is used to identify the denoised speech.The simulation results show that the recognition rate of the hybrid model is higher.Finally,the proposed algorithm in this paper is applied to the robotic voice control system,and the logistics sorting simulation system is built and run.The test results show that the software developed in this paper can realize the control and sorting of different types of goods by voice-command control manipulator,which proves that the proposed algorithm can be applied to the robot voice-command control system.