Motion Trajectory Planning of Hexapod Robot Based on Improved SLAM Algorithm
Existing control algorithms have the shortages of large trajectory control deviation and poor collision avoidance ability,this paper researches the control process of hexapod robot,and proposes a control scheme based on simultaneous localization and map-ping(SLAM)algorithm.The Denavit-Hartenberg(D-H)model is studied,and the spatial motion process of the hexapod robot is an-alyzed.a high definition(HD)camera is used to acquire the images of the job site and realize the coordinate conversion of the HD camera at the same time.an inertial measurement unit(IMU)unit is used to improve the stability of the SLAM algorithm model and carry out the calibration of errors and parameters,the local binarization model is based on extracting the features from the field ima-ges.In the image feature set training,the convolutional neural network(CNN)network model is used to improve the data training a-bility of the SLAM algorithm model,and according to the maximum discount reward value after interacting with the field environ-ment,the gait stability of the robot and collision avoidance effect of the local area are improved.Experimental results show that the improved SLAM algorithm can achieve a trajectory path optimization in global scope,with a path time of only 35.4 s,and only one of ten collision avoidance tests occurs the improved SLAM algorithm is superior to other collision control algorithms.
bionic hexapod robotcollision avoidancelocal binarizationmaximum discount reward value
NETL
NSTL
万方数据
