The Research On Autonomous Montion Positioning For Robots Based On Gyroscope

Autonomous mobile robot is more and more widely applied in military andcivilian, regardless of what kind of functions, accurating to arrive destination isthe basis of achieving goals so for a robot autonomous motion planning in acomplex environment put forward higher requirements.This paper studies the keytechnologies of the robot motion planning-positioning,which is determined therelative position of the robot with the starting point and moving objects in realtime and lay the foundation for the next motion planning.The robot positioning system designed by this paper is divided into foursteps:positioning data acquisition,positioning information decoding, positioningerror elimination and trajectory simulation.First, the motion sensor based on theMPU-6050chip real-time acquisition robot motion data: triaxial accelerometerand triaxial angular velocity to obtain positioning data;Then converting thethree-axis angular velocity to robot motion attitude Angle based on thequaternion algorithm to get the positioning information; Due to a variety ofexternal factors, the positioning information gathered by motion sensor exists theinevitable drift error. Handling robot motion planning data through Kalman filtertheory of optimal estimation, the error on the path planning effects areminimized.This paper use the method of uniform motion in a straight line of robots anduniform circular motion positioning data collected by calculating and noisereduction processing,then based on the data processed, we simulated robottrajectory. This paper put forward innovativily by using gyroscope collectedthree shaft angular velocity and axial acceleration simulation robot trajectoryalgorithm, the robot respectively through two kinds of sports,convert positioningdata into motion reference frame coordinate values,by using the coordinatevalues draw the robot trajectory. Compared with the error of data and data afterextended kalman filtering respectively simulated robot trajectory, it can be seenafter filtering effect is close to the ideal trajectory,proved in this paper, thedesign of real-time positioning system can determine and predict the movementtrajectory of robot, and limit error in the controllable range.