基于概率神经网络图像识别的移动机器人控制研究
|
摘要
为实现移动物流机器人轨迹自动控制,本文研究了基于概率神经网络图像识别的移动物流机器人控制系统,设计了基于概率网络的图像采集和识别处理模块,研究了移动物流机器人自主运动建模、控制的关键问题.通过搭建基于概率神经网络的图像采集和识别处理模块、移动物流机器人轨迹控制数学模型,在该模型基础上设计移动机器人基于图像识别结果的路径控制算法.利用仿真软件MATLAB的Robotic tool工具箱,进行移动机器人轨迹控制仿真.仿真结果证明,本文的路径控制方案可以让移动机器人根据图像识别结果很好的完成预定路径控制,在强度为0. 4椒盐噪声下路径控制正确率约为98%,为实现噪声环境下基于视觉移动机器人轨迹自主控制提供了理论依据和应用基础.
In order to realize the automatic trajectory control of the mobile logistics robot,this paper studies the mobile logistics robot control system based on image recognition results using the probabilistic neural network. In this paper,we design the image acquisition and recognition processing module based on the probabilistic neural network and studies the key problems of the modeling and control of the autonomous motion of the mobile logistics robot. By setting up the image acquisition and recognition processing module based on probabilistic neural network and the mathematical model of the trajectory control of the mobile logistics robot,a path control algorithm based on the image recognition results of this kind of mobile robot is designed. The Robotic tool toolbox of simulation software MATLAB is used to simulate the control trajectory of mobile robot. The simulation results show that the path control scheme in this paper can make the mobile robot achieve the desired path tracking successfully according to the image recognition results. Under the condition of 0. 4 salt and pepper noise,the correct rate of path control is 98%,and the theoretical basis and application foundation are provided for the realization of autonomous control of the trajectory of mobile robot based on vision under noise environment.
In order to realize the automatic trajectory control of the mobile logistics robot,this paper studies the mobile logistics robot control system based on image recognition results using the probabilistic neural network. In this paper,we design the image acquisition and recognition processing module based on the probabilistic neural network and studies the key problems of the modeling and control of the autonomous motion of the mobile logistics robot. By setting up the image acquisition and recognition processing module based on probabilistic neural network and the mathematical model of the trajectory control of the mobile logistics robot,a path control algorithm based on the image recognition results of this kind of mobile robot is designed. The Robotic tool toolbox of simulation software MATLAB is used to simulate the control trajectory of mobile robot. The simulation results show that the path control scheme in this paper can make the mobile robot achieve the desired path tracking successfully according to the image recognition results. Under the condition of 0. 4 salt and pepper noise,the correct rate of path control is 98%,and the theoretical basis and application foundation are provided for the realization of autonomous control of the trajectory of mobile robot based on vision under noise environment.
引文
[1]Wang Lin-hua.Design of magnetic AGV automatic car navigation control system[D].Changsha:Changsha University of Science&Technology,2013.
[2]Tlale N S.Fuzzy logic controller with slip detection behaviour for mecanum-wheeled AGV[J].Robotica,2005,23(4):455-456.
[3]Baber J,et al.Cooperative autonomous driving:intelligent vehicles sharing city roads[J].IEEE Robotics&Automation M agazine,2005,12(1):44-49.
[4]Liu Wei.Research on automatic guided vehicle control system[D].Changchun:Jilin University,2015.
[5]Written by Peter Corke,translated by Liu Rong.Robotics,machine vision and control-M ATLAB algorithm foundation[M].Beijing:Tsinghua University Press,2016.
[6]Chen Ming,et al.Principles and examples of MATLAB neural network[M].Bejing:Tsinghua University Press,2016.
[7]Wang Xiao-chuan,Shi Feng,Yu Lei,et al.43 case studies of MAT-LAB neural network[M].Beijing:Beihang University Press,2013.
[8]Yuan Xin.Research and implementation of engine fault diagnosis system based on neural network[D].Dalian:Dalian University of Technology,2007.
[9]Sprecht D F.Probabilistic neural networks for classification,mapping and associative memory[J].IEEE International Conference on Neural Networks,1988,1(1):525-532.
[10]George D,Lehrach W,Kansky K,et al.A generative vision model that trains with high data efficiency and breaks text-based CAPT-CHAs[J].Science,2017,358(6368):2612.
[11]Itoh T,Takeda K,Itakura F.Aeoustic analysis and recognition of whispered spoeeh[C].Orlando,Florida,USA,2002:389-392.
[12]Weisbin C R,Rodriguez G.NASA robotics research for planetary surface exploration[J].IEEE Robotics&Automation M agazine,2002,7(4):25-34.
[13]Francoismichaud,Dominicletourneau,Martinarsenault,et al.Multimodal locomotion robotic platform using leg-track-wheel articulations[J].Autonomous Robots,2005,(18):137-156.
[14]Contreras-Cruz M A,Ayala-Ramirez V,Hernandez-Belmonte U H.M obile robot path planning using artificial bee colony and evolutionary programming[J].Applied Soft Computing Journal,2015,30:319-328.
[1]王琳华.磁导式AGV自动导航车控制系统的设计[D].长沙:长沙理工大学,2013.
[4]刘巍.自动导引小车(AGV)控制系统研究[D].长春:吉林大学,2015.
[5]Peter Corke著,刘荣等译.机器人学、机器视觉与控制-MATLAB算法基础[M].北京:清华大学出版社,2016.
[6]陈明,等.MATLAB神经网络原理与实例精讲[M].北京:清华大学出版社,2016.
[7]王小川,史峰,郁磊,等.MATLAB神经网络43个案例分析[M].北京:北京航空航天大学出版社,2013.
[8]袁鑫.基于神经网络的发动机故障诊断系统的研究与实现[D].大连:大连理工大学,2007.
[2]Tlale N S.Fuzzy logic controller with slip detection behaviour for mecanum-wheeled AGV[J].Robotica,2005,23(4):455-456.
[3]Baber J,et al.Cooperative autonomous driving:intelligent vehicles sharing city roads[J].IEEE Robotics&Automation M agazine,2005,12(1):44-49.
[4]Liu Wei.Research on automatic guided vehicle control system[D].Changchun:Jilin University,2015.
[5]Written by Peter Corke,translated by Liu Rong.Robotics,machine vision and control-M ATLAB algorithm foundation[M].Beijing:Tsinghua University Press,2016.
[6]Chen Ming,et al.Principles and examples of MATLAB neural network[M].Bejing:Tsinghua University Press,2016.
[7]Wang Xiao-chuan,Shi Feng,Yu Lei,et al.43 case studies of MAT-LAB neural network[M].Beijing:Beihang University Press,2013.
[8]Yuan Xin.Research and implementation of engine fault diagnosis system based on neural network[D].Dalian:Dalian University of Technology,2007.
[9]Sprecht D F.Probabilistic neural networks for classification,mapping and associative memory[J].IEEE International Conference on Neural Networks,1988,1(1):525-532.
[10]George D,Lehrach W,Kansky K,et al.A generative vision model that trains with high data efficiency and breaks text-based CAPT-CHAs[J].Science,2017,358(6368):2612.
[11]Itoh T,Takeda K,Itakura F.Aeoustic analysis and recognition of whispered spoeeh[C].Orlando,Florida,USA,2002:389-392.
[12]Weisbin C R,Rodriguez G.NASA robotics research for planetary surface exploration[J].IEEE Robotics&Automation M agazine,2002,7(4):25-34.
[13]Francoismichaud,Dominicletourneau,Martinarsenault,et al.Multimodal locomotion robotic platform using leg-track-wheel articulations[J].Autonomous Robots,2005,(18):137-156.
[14]Contreras-Cruz M A,Ayala-Ramirez V,Hernandez-Belmonte U H.M obile robot path planning using artificial bee colony and evolutionary programming[J].Applied Soft Computing Journal,2015,30:319-328.
[1]王琳华.磁导式AGV自动导航车控制系统的设计[D].长沙:长沙理工大学,2013.
[4]刘巍.自动导引小车(AGV)控制系统研究[D].长春:吉林大学,2015.
[5]Peter Corke著,刘荣等译.机器人学、机器视觉与控制-MATLAB算法基础[M].北京:清华大学出版社,2016.
[6]陈明,等.MATLAB神经网络原理与实例精讲[M].北京:清华大学出版社,2016.
[7]王小川,史峰,郁磊,等.MATLAB神经网络43个案例分析[M].北京:北京航空航天大学出版社,2013.
[8]袁鑫.基于神经网络的发动机故障诊断系统的研究与实现[D].大连:大连理工大学,2007.