The Correlative Positioning Error Compensation of the Vision System and the Robot Mechanism Based on BP Neural Network

Abstract:

Error analysis and compensation of positioning is one of the basic problems for robot control. There are some limitations to study the error compensation either from the perspective of vision system or mechanism of the robot. Therefore, this paper studied the error compensation from both aspects with BP neural network model. Firstly, aiming at the depth values obtained by the vision system, a depth error database was established and based on BP neural network the depth error was effectively compensated within 5mm in MATLAB simulation. Secondly, aiming at the mechanism of a robot prototype the error compensation simulation based on BP neural network in MATLAB was carried out, and the simulation result showed that the error was within ±1mm after compensation. Thirdly, the correlative positioning error was defined and its error compensation simulation was carried out in MATLAB with the BP neural networks of the vision system and the mechanism; the simulation results showed that the correlative positioning error was approximately zero in this study. Finally, the error compensation of the robot mechanism based on the BP neural network was carried out on the robot prototype HNrobot2, and the experiment results showed that the positioning error was by 60% less, which was within 5mm.

• Periodical: Key Engineering Materials (Volume 621)
• Pages: 513-518
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.621.513
• Online since: August 2014
• Authors: Hai Xin Zou*, Xiang Jun Zou, Jun Tao Xiong, Min Ye
• Keywords: BP Neural Network (BPNN), Correlative Positioning Error, Error Compensation, Robot, Simulation

References

[1] J.T. Xiong: Study on Visual Position of Litchi Picking Manipulator in Complex Environment (Ph.D., South China Agricultural University, China 2012), pp.85-91 (In Chinese).

[2] X.L. Lv: Study on Vision System Based on the Four Degrees of Freedom Tomato Picking Robot (Ph.D., Shenyang Agricultural University, China 2008), pp.87-96 (In Chinese).

[3] J. Lu: The Association Positioning of Picking-machine and Vision Based on Virtual Design (Master, South China Agricultural University, China 2009), pp.57-66 (In Chinese).

[4] Y. Wang, F. Gao and Q.H. Yang: Modem Manufacturing Engineering, (2011) No. 8, pp.57-61. (In Chinese).

[5] Y. Wang: Research on Motion Planning and Control System of the Cucumber Picking Robot (Ph.D., Zhejiang University of Technology, China 2010), pp.62-70 (In Chinese).

[6] K O Ker, R.C.S.I., et al.: Robotics and Autonomous Systems, Vol. 49(2004), pp.227-234.

[7] K O Ker, R.C.S.I.: Information Sciences. Vol. 222(2013), pp.528-543.

[8] Zou, X., H. Zou and J. Lu: Machine Vision and Applications, Vol. 23(2012), pp.43-63.