Research On Motion Control Of Four-wheel Omni-directional Mobile Robot

With the development of science, robot has been used in serious fields such as industry, agriculture,medical instrument and so on.In a Seminar on major challenges in the field of artificial intelligence held in Tokyo, the researchers on manufacturing and training robot football game to promote related research are discussed, under some scholars actively advocating and with the support of SON Y company, the RoboCup federation was set up, and in1996, the exhibition game was successfully held.World Cup soccer robot competition has been held every year since1997, and the event has greatly promoted robot researches in various countries. RoboCup received extensive attention of universities, and became an important hot topic.RoboCup related to precision instrument design, machine vision, robot motion control, multi-sensor information fusion, distributed network, and many other fields, this paper carried out the robot precision instrument design, the research of motion control, etc.At present, the study on the robot control, mainly refer to the optimization of the mobile platform, upgrade hardware performance, improved control algorithm and so on.On robot four-wheel platforms, the omni-directional wheel with its flexibility take great advantages, wheel structures widely adopt three or four wheel structure; Processor mostly adopts high-performance processor, such as DSP, arm, industrial control machine, etc; Control algorithm mainly improves the performance of the robot control algorithm reduces the complexity of the algorithm, etc.In this paper, the work is as follows:First, control robot to track on speed trajectory. The robot kinematics equation is established, and real time compensate speed value of each motor given, makes the robot to achieve better trajectory tracking.Second, BP neural network speed compensator design of robot. Using BP neural network and strong memory capacity and nonlinear Function approximation ability, according to the relationship between design speed compensator error and the compensation amount, design and simulate the BP neural network speed compensator.Third, under a state of disturb, the introduce the sliding mode control algorithm for trajectory tracking of robot manipulators. The sliding mode control for the dynamic model of the robot is established, and comparing with the PD controller.